article
stringlengths 0
745k
| abstract
stringlengths 0
92.7k
|
|---|---|
the monte carlo method based on the metropolis algorithm is the most successful and influential stochastic algorithm of the 20th century and has been used in variety of applications not limited to physics @xcite .
the method is powerful in an exhaustive search of highly multi - dimensional phase space , and , hence , has been routinely used to calculate the thermal averaging of statistical physics . aside from the applications to statistical physics , the metropolis algorithm has been used as a vehicle for global optimization , that is , a task to search for the lowest minimum point in a rugged landscapes in a high dimension .
in fact , the simulated annealing ( sa ) @xcite based on the metropolis algorithm is the oldest metaheuristics in global optimization . in global optimization
, a good balance between a global search ( exploration ) and a local search ( exploitation ) is necessary .
since the metropolis algorithm has only the ability to perform a global search , it is usually necessary to augment sa using a more traditional local optimization method to handle realistic problems @xcite .
later , a method which combines the metropolis algorithm and the gradient - based local search algorithm was proposed .
that is the monte carlo minimization ( mcm ) of li and sherga @xcite , and the basin - hopping ( bh ) of wales and doye @xcite .
wales and doye @xcite , for example , used their bh algorithm to study the lowest - energy structures of lennard - jones clusters that consist of up to 110 atoms successfully . in this paper , we propose a new variant of the basin - hopping(bh ) algorithm .
in contrast to another variant of the bh by leary and doye @xcite where only the down - hill search is allowed , we borrow the concept of extremal optimization @xcite or thermal cycling @xcite , and introduce the process of _ jumping _ to enhance the search of a rugged energy landscape .
the basin hopping ( bh ) algorithm ( which is also called the monte carlo plus minimization , mcm ) uses hopping due to monte carlo random walks to explore the phase space , and gradient - based local optimization to locate the local minimum or _ basin _ plus metropolis criterion to accept or reject the move .
since the energy landscape explored by the usual monte carlo move and immediate relaxation to a nearby basin of attractor using the gradient method in bh looks like hopping among basin of attractors , the algorithm is termed `` basin hopping '' . in order to enhance the exploration , the monte carlo move in bh
consists of a simultaneous displacement of all particles in the cluster in contrast to the usual monte carlo method in classical statistical mechanics @xcite where usually a randomly chosen single particle is displaced .
[ fig : energy1 ] the bh has been extensively tested @xcite for the simplest benchmark problem for the lennard jones clusters lj@xmath1 @xcite where the total energy of @xmath2-atom cluster is given by @xmath3 where @xmath4 is the distance between two particles @xmath5 and @xmath6 within the cluster . even for this simple lennard - jones problem
, the bh clarified the extremely complex energy landscape for several clusters @xcite , and it even failed to locate the lowest - energy structures for several special sizes .
for example , the bh could not locate the lowest minimum of lj@xmath7 , lj@xmath8 , and lj@xmath9 easily @xcite when an unbiased search that starts from a completely random initial configuration of clusters is used .
in fact , in order to find out the lowest - energy structure of lj@xmath10 , wales and doye @xcite had to use _ seeding _ to feed the bh the lowest - energy structure of smaller lj@xmath11 or larger lj@xmath12 clusters .
the unbiased search has an extremely low probability ( 4% ) of hitting the lowest - energy structure even when large numbers ( 100 times ) of a fairly long run ( 5000 step ) were executed .
the same problem occurred for lj@xmath9 , and less severely in lj@xmath13 , lj@xmath14 and lj@xmath15 .
difficulty occurs when the energy landscape consists of several large valleys ( funnels ) instead of one , and the funnel corresponding to the lowest - energy minimum is narrow and separated from other funnels by a high barrier @xcite .
the monte carlo trial move plus local minimization and acceptance / rejection using the metropolis criterion is less powerful and time - consuming to overcome such a large barrier ( fig .
recently , a new variant of the basin - hopping algorithm called the `` monotonic sequence basin - hopping algorithm '' ( msbh ) was proposed by leary and doye @xcite .
their algorithm is essentially the bh at temperature @xmath16 , which allows only downhill moves .
when the search is stacked at a local minimum , the program restarts from a new random configuration .
therefore , msbh @xcite is essentially the multi - start strategy of a greedy search .
naturally , the msbh algorithm seems less powerful @xcite than the original bh algorithm because it is not equipped with any mechanism to cross the barrier .
one of the present authors @xcite suggested another extension of the bh by introducing the concept of `` extremal optimization '' ( eo ) proposed by boettcher and percus @xcite . in this
eo - based basin - hopping ( eobh ) algorithm , the monte carlo move of only one less - bounded particle within the cluster is attempted , and every move is accepted without using the metropolis criterion .
therefore this eobh is essentially the bh at @xmath0 .
although the eobh can achieve the crossing of high barriers in contrast to the msbh , it is again less powerful than the original bh because its ability in terms of a local search is less effective though the method is proven to be useful to enumerate all the low - energy structures @xcite . the lesson learned from these two previous exercises is that the inclusion of the high - temperature monte carlo move at @xmath0 will enhance the ability of a global search ( exploration ) , but the original metropolis criterion of the monte carlo move should be retained to maintain the efficiency of a local search ( exploitation ) .
actually , such a prescription of re - heating or thermal treatment at high temperatures has been repeatedly proposed in the application of the simulated annealing ( sa ) @xcite .
for example , mbius _ et al _ , introduced `` thermal cycling '' and ingber @xcite introduced `` reannealing '' in sa . in order to enhance the ability of crossing the high barrier in the bh , we have introduced a re - heating process called `` jumping '' as shown in fig .
fig : flow2 ] jumping is the monte carlo move without relaxation ( local minimization ) at @xmath0 , and , hence , it is always accepted .
when the usual monte carlo moves are rejected max times , the system is judged to be trapped at the local minimum .
then the temperature is raised to @xmath0 , and the monte carlo moves ( jumpings ) are executed jmp times in the search space to allow for the system to escape from the local minimum ( fig .
this move is always accepted irrespective of the increase in the energy because of @xmath0 and is called jumping here .
the parameter max is used to detect the entrapment .
the parameter jmp is used to try to climb up the barrier several times . if the jmp is too large , the algorithm is nothing more than simple random multi - start strategy .
it is essential to keep the partial memory of the previous state @xcite .
so , the jmp should not be too large .
we will call this version of bh the `` basin - hopping with occasional jumping '' ( bhoj ) . [
fig : energy2 ] now , the exploration of phase space in our bhoj proceeds as follows ( fig .
3 ) : the rugged energy landscape is transformed into successive steps of the basin by local minimization .
the downhill move is simply the descending of the stairs by hopping .
there are , however , two kinds of uphill moves .
one is by climbing the stairs by hopping using the metropolis criterion , which is costly and may not be effective to climb up the high barrier because the moves with a large energy difference are rejected by the metropolis criterion .
another move is jumping , which does not use local minimization and is always accepted , so the uphill moves by jumping do not use stairs and are simply along the surface of the hill ( fig .
3 ) . this jumping must be an efficient way to escape from a local minimum ( valley ) and to explore the next basin of the valley when it is separated by high barriers .
in order to test the performance of our modification of the bh with occasional jumping ( bhoj ) , we have calculated the lowest - energy structure of the lennard - jones clusters of a particular size , lj@xmath17 , lj@xmath7 , lj@xmath8 , and lj@xmath9 for which the original bh is not effective .
we conducted a 100 independent unbiased search starting from 100 random initial structures with the maximum of 5000 iterations which do not include the jumping process .
we also performed the same experiment using the original bh @xcite with the same random initial structures .
the temperature is fixed at @xmath18 .
two additional parameters max and jmp for the bhoj are arbitrarily fixed to max=10 and jmp=3 or 5 or 7
. table 1 gives the success rates of 100 unbiased searches .
our bhoj could successfully reproduce the lowest - energy structures in the literature @xcite .
it is apparent that our bhoj performs in general better than the original bh .
the success rate increased twice to five times from the original bh to the bhoj .
thus , the ability of exploration is in fact enhanced by the introduction of jumping processes . for the sake of comparison
, we also showed the results of msbh of leary @xcite and the parallel fast annealing ( pfa ) of cai _ et al . _
@xcite . .
success rates of original bh @xcite , msbh @xcite , pfa @xcite and our bhoj for selected lennard - jones clusters which are notoriously difficult cases to find out the lowest - energy structure . in msbh
leary conducted experiments 1000 times while we conducted experiments 100 times for original bh and bhoj . [ cols="^,>,>,>,>,>",options="header " , ] table 2 shows the results for lj@xmath15 and lj@xmath19 . for lj@xmath15 we could confirm the new lowest - energy structure found by wales and doye @xcite using our bhoj .
we could also successfully confirm the lowest energy -1125.493794 cited in @xcite found by leary for lj@xmath20 which is lower than the previous record -1125.304876 found by hartke @xcite .
the new lowest - energies -1132.669966 for lj@xmath21 and -1139.455696 for lj@xmath22 found by hartke @xcite were also successfully located by our bhoj though the success rates of these three cases were very low .
in this paper we proposed a way to improve the performance of the basin hopping ( bh ) algorithm by introducing the jumping in addition to the hopping .
we call this new algorithm as the basin - hopping with occasional jumping ( bhoj ) .
the jumping is a process of heating the system and raising the temperature to infinitely high which is attempted when the trajectory in the phase space is trapped at a local minimum . by jumping
, the trajectory can climb up high barriers and can explore the next valley .
thus the exploratory ability of the algorithm is enhanced .
experiments on benchmark problem of the lennard - jones clusters , in particular , for notorious difficult sizes of 75 to 77 particles lj@xmath7 , of 98 particles lj@xmath8 , and of 102 to 104 particles lj@xmath9 reveal that the proposed bhoj is really superior to the original bh .
this jumping is easy to implement , and consumes very little cpu resources .
any adaptive or scheduled jumping could be easily incorporated .
the bhoj with jumping will be helpful to search for the lowest - energy structures of larger clusters and more complex clusters with many body forces .
one of the author ( m.i . ) would like to thank department of physics , tokyo metropolitan university for providing him a visiting fellowship .
22 i. beichl and f. sullivan , computing in science and engineering * 2 * ( 2000 ) 65 .
s. kirkpatrick , c.d .
gelatt , m.p .
vecchi , science * 220 * ( 1983 ) 671 .
wille , chem .
* 133 * ( 1987 ) 405 .
z. li , h.a .
scheraga , proc .
* 84 * ( 1987 ) 6611 .
wales , j.p.k .
doye , j. phys .
chem . a * 101 * ( 1997 ) 5111 .
leary and j.p.k .
doye , phys .
e * 60 * ( 1999 ) r6320 .
leary , j. global .
* 18 * ( 2000 ) 367 .
p. boettcher and a. percus , artificial intelligence * 119 * ( 2000 ) 275 .
a.mbius , a. neklioudov , a. daz - snchez , k.h .
hoffmann , a. fachat , m. schreiber , phys .
* 79 * ( 1997 ) 4297 .
d. frenkel , b. smit , _ understanding molecular simulation from algorithm to application _ , academic press , 1986 .
hoare , p. pal , adv .
* 20 * ( 1971 ) 161 .
northby , j. chem . phys .
* 87 * ( 1987 ) 6166 .
deaven , n. tit , j.r .
morris , k.m . ho , chem .
* 256 * ( 1996 ) 195 .
niesse , h.r .
mayne , j. chem .
* 105 * ( 1996 ) 8428 .
doye , m.a .
miller , d.j .
wales , j. chem . phys .
* 111 * ( 1999 ) 8417 .
m. iwamatsu , proc .
2002 ieee world congr .
pp.1180 , ieee press , 2002 .
l. ingber , mathl .
modelling * 12 * ( 1989 ) 967 .
wales , program `` gmin '' , http://www-wales.ch.cam.ac.uk /software.html .
wales , `` cambridge cluster database '' , + http://www-doye.ch.cam.ac.uk/jon/structures/lj /tables.150.html ) .
w. cai , h. jiang , x. shao , j. chem .
* 42 * ( 2002 ) 1099 .
m.d . wolf and u. landmann ,
a * 102 * ( 1998 ) 6129 .
j. pillardy , a. liwo , h.a .
scheraga , j. phys .
a * 103 * ( 1999 ) 9370 .
g. hartke , j. comput . chem . * 20 * ( 1999 ) 1752 .
i. rata , a.a .
shvartsburg , m. horoi , t. frauenheim , k.w.m .
siu , k.a .
jackson , phys .
* 85 * ( 2000 ) 546 .
j. lee , i - h .
lee , j. lee , phys .
* 91 * ( 2003 ) 080201 .
d. romero , c. barrn , s. gmez , comp .
* 123 * ( 1999 ) 87 . c. barrn , http://www.vcl.uh.edu/@xmath23cbarron/lj_ cluster /ljpottable.html
( currently down ) . | basin - hopping ( bh ) or monte - carlo minimization ( mcm ) is so far the most reliable algorithms in chemical physics to search for the lowest - energy structure of atomic clusters and macromolecular systems .
bh transforms the complex energy landscape into a collection of basins , and explores them by hopping , which is achieved by random monte carlo moves and acceptance / rejection using the metropolis criterion . in this report
, we introduce the jumping process in addition to the hopping process in bh . jumping
are invoked when the hopping stagnates by reaching the local optima , and are achieved using the monte carlo move at the temperature @xmath0 without rejection .
our basin - hopping with occasional jumping ( bhoj ) algorithm is applied to the lennard - jones clusters of several notoriously difficult sizes .
it was found that the probability of locating the true global optima using bhoj is significantly higher than the original bh .
basin - hopping , lowest - energy structure , lennard - jones cluster 02.60.pn , 02.70.tt , 36.40.mr |
in mesoscopic samples there is a competition between a triangle configuration of the vortex lattice as being the lowest energy configuration in bulk material ( and films ) and the boundary which tries to impose its geometry on the vortex lattice .
for example a circular geometry will favour vortices situated on a ring near the boundary and only far away from the boundary its influence diminishes and the triangular lattice may reappear .
therefore , it is expected that different geometries will favour different arrangements of vortices and will make certain vortex configurations more stable than others
. in small systems vortices may overlap so strongly that it is more favourable to form one big giant vortex .
the latter will preferably have a circular geometry . as a consequence
it is expected that the giant to multivortex transition will be strongly influenced by the geometry of the boundary as will be also the stability of the giant vortex configuration .
these issues , the dependence of the stability of the giant vortex configuration and of the different multivortex configurations on the geometry of the sample will be investigated in the present paper . as an example
, we will compare the most important geometries : the circular disk , the square and the triangle .
mesoscopic ( circular ) disks and rings have been the most popular in this respect .
experimentally , the magnetization of superconducting disks has been measured as a function of the externally applied magnetic field @xcite .
several transitions between different superconducting states were found and the magnetization depends sensitively on size and temperature .
the main part of the theoretical studies covered disks @xcite and rings @xcite of zero thickness . in this case
one can neglect the magnetic field induced by the supercurrents and one assumes that the total magnetic field equals the external applied magnetic field , which is uniform .
a limited number of studies considered disks @xcite and rings @xcite with _
finite _ thickness .
then , the finite thickness of the disks influences the magnetic field profile and it is necessary to take into account the demagnetization effects
. often only the ( circular symmetric ) giant vortex states or the superconducting / normal transition were investigated . even in type - i superconductors
multivortex states in disks @xcite and rings @xcite were predicted .
it was found that if the disk or the ring is large enough , the giant vortex nucleates into a multivortex state in which the vortices are situated on a ring . in a ring geometry , we found that breaking the circular symmetry through a non - central location of the hole favors the multivortex state @xcite .
this means that by changing the geometry , the giant vortex state transits into a multivortex state .
mesoscopic superconductors with non - circular geometries have attracted less attention .
et al _ @xcite measured the superconducting / normal transition in superconducting lines , squares and square rings using resistance measurements .
et al _ @xcite calculated the @xmath0 phase diagram for a square with zero thickness in the framework of the linearized ginzburg - landau theory , which is only valid near the superconducting / normal boundary .
they compared their results with the @xmath0 phase boundary obtained from resistance measurements .
et al _ @xcite studied square loops with leads attached to it and found inhomogeneous cooper - pair distributions in the loop with enhancements near the corners of the square loop .
et al _ @xcite calculated the nucleation field as a function of the sample area for disks , squares and triangles with zero thickness .
et al _ @xcite computed the superconducting / normal transition for mesoscopic disks and squares of zero thickness . for macroscopic squares , the magnetic field distribution and the flux penetration
are investigated in detail by experimental observations using the magneto - optical faraday effect and by first - principles calculations which describe the superconductor as a nonlinear anisotropic conductor @xcite . in the latter case
the penetration of the magnetic field occurs continuously . in macroscopic samples the penetration of individual fluxoids is not so important in the theoretical description of the magnetic response of the superconductor , but it turns out to be essential in our mesoscopic samples .
recently , aftalion and du @xcite studied cylindrical square shaped superconductors within the ginzburg - landau theory .
et al _ investigated the vortex entry and the nucleation of anti - vortices in infinite thin superconducting squares @xcite and triangles @xcite using the linearized ginzburg - landau theory . within this linear theory
they studied the superconducting / normal transition and they found near this transition the nucleation of multivortices , anti - vortices and combination of these two instead of the expected surface superconductivity .
they also calculated the @xmath0 phase diagrams for the square and the triangle .
recently , bona and kabanov @xcite studied thin superconducting squares using the nonlinear ginzburg - landau theory in the @xmath1 limit . within this non - linear theory
they showed that the vortex - anti - vortex configuration becomes unstable when moving away from the superconducting / normal transition . in the present paper
we consider superconductors of finite thickness and study also the vortex configurations deep inside the superconducting state , i.e. far from the superconducting / normal boundary .
our main focus will be on the influence of the geometry of the superconductor on the vortex configuration and its stability .
our theoretical analysis is based on a full self - consistent numerical solution of the coupled nonlinear ginzburg - landau equations for arbitrary value of @xmath2 .
no a priori shape or arrangement of the vortex configuration is assumed .
the magnetic field profile near and in the superconductor is obtained self - consistently , and therefore the full demagnetization effect is included in our approach .
the paper is organized as follows : in sec .
ii we describe the theoretical formalism .
our results are presented in sec .
we calculate and compare the free energy and the magnetization for disks , squares and triangles with the same surface area .
next , we make a distinction between multivortex states and giant vortex states and we investigate the influence of the sample geometry on the vortex lattice .
we also calculate the magnetic field range over which the vortex states with vorticity @xmath3 are stable in disks , squares and triangles . the magnetic field distribution and the current density are studied and the @xmath0 phase diagram is obtained .
finally , we summarized our results in sec .
in the present paper , we consider thin superconducting samples having the same volume but with different geometry which are immersed in an insulating medium in the presence of a perpendicular uniform magnetic field @xmath4 . to solve this problem
we follow the numerical approach of schweigert and peeters @xcite . as for thin disks
@xmath5 it is allowed to average the gl equations over the disk thickness for samples of arbitrary geometry .
using dimensionless variables and the london gauge @xmath6 for the vector potential @xmath7 , we write the system of gl equations in the following form @xmath8 @xmath9 where @xmath10 is the density of superconducting current .
the superconducting wavefunction satisfies the boundary conditions @xmath11 normal to the sample surface and @xmath12 far away from the superconductor . here
the distance is measured in units of the coherence length @xmath13 , the vector potential in @xmath14 _ h-.2em.2em_@xmath15 , and the magnetic field in @xmath16 _ h-.2em.2em_@xmath17 .
the superconductor is placed in the @xmath18 plane , the external magnetic field is directed along the @xmath19 axis , and the indices 2d , 3d refer to two- and three - dimensional operators , respectively . to solve the system of eqs .
1(a , b ) , we generalized the approach of ref .
@xcite for circular disks to superconductors with an arbitrary flat geometry .
we apply a finite - difference representation for the order parameter and the vector potential on an uniform cartesian space grid ( x , y ) , with typically @xmath20 grid points for the area of the superconductor , and use the link variable approach @xcite , and an iteration procedure based on the gauss - seidel technique to find @xmath21 .
the vector potential is obtained with the fast fourier transform technique where we set @xmath22 at the boundary of a box with a larger space grid of size typically 4 times the superconductor area .
for circular configurations like disks the giant vortex state is characterized by the total angular momentum @xmath3 through @xmath23 where @xmath24 and @xmath25 are the cylindrical coordinates .
@xmath3 is the winding number and gives the vorticity of the system . due to the non - linearity of the gl equations an arbitrary superconducting state is generally a mixture of different angular harmonics @xmath26 even in axially symmetric systems
nevertheless , we can introduce an analog to the total angular momentum @xmath3 which is still a good quantum number and which is in fact nothing else then the number of vortices in the system . for non - axially symmetric systems there exists no axially symmetric giant vortex states and
hence the superconducting state is always a mixture of different angular harmonics .
the vorticity @xmath3 of a particular superconducting sample can be calculated by considering the phase @xmath27 of the order parameter along a closed loop near the boundary of the sample , where the total phase difference is always @xmath28 . in non - axially
symmetric systems three possible vortex states exist : ( i ) a multivortex state which contains separate vortices , ( ii ) a superconducting state which contains one giant vortex in the center , and ( iii ) a state which is a mixture of both : a giant vortex in the center which is surrounded by single vortices . the giant vortex is not necessary circular symmetric as in the case of a circular disk , but it may be deformed due to the specific shape of the sample boundary . to find the different vortex configurations , which include the metastable states , we search for the steady - state solutions of eqs .
1(a , b ) starting from different randomly generated initial conditions .
then we increase / decrease slowly the magnetic field and recalculate each time the exact vortex structure .
we do this for each vortex configuration in a magnetic field range where the number of vortices stays the same . by comparing the dimensionless gibbs free energies of the different vortex configurations
@xmath29 d% \overrightarrow{r } , \eqnum{2}\ ] ] where integration is performed over the sample volume @xmath30 and @xmath31 is the vector potential of the uniform magnetic field , we find the ground state .
the dimensionless magnetization , which is a direct measure of the expelled magnetic field from the sample , is defined as @xmath32 where @xmath4 is the applied magnetic field .
@xmath33 is the magnetic field averaged over the sample and @xmath34 @xmath35 .
the temperature is indirectly included in @xmath13 , @xmath36 , @xmath37 , whose temperature dependence is given by @xmath38 where @xmath39 is the critical temperature at zero magnetic field .
we will only explicitly insert temperature if we consider the @xmath0 phase diagrams , while the other calculations are for a certain fixed temperature .
notice further that the ginzburg - landau parameter @xmath40 is independent of the temperature .
as a typical example , we consider superconducting disks , squares and triangles with the same surface area @xmath41 , the same finite thickness @xmath42 and the same ginzburg - landau parameter @xmath43 which is typical for al @xcite . thus the superconducting disk has a radius @xmath44 , the square has a width @xmath45 and the triangle has a width @xmath46 . in the first step
, we will compare the free energy and the magnetization for the three geometries . in figs .
[ emag](a , b ) the free energy and the magnetization are shown for the disk as a function of the applied magnetic field , in figs .
[ emag](c , d ) for the square and in figs .
[ emag](e , f ) for the triangle . in figs .
[ emag](a , c , e ) the free energy of the different giant vortex states is given by solid curves and the free energy of the multivortex states by dashed curves .
the open circles indicate the transition from multivortex state to giant vortex state at the transition fields @xmath47 .
this transition is of second order @xcite
[ emag](b , d , f ) show the magnetization of the different @xmath3-states with thin curves and the ground state is indicated by the thick curve .
there exists vortex states with vorticity up to @xmath48 for the disk and the square and up to 13 for the triangle .
the superconducting state is destroyed at @xmath49 for the disk , at @xmath50 for the square and @xmath51 for the triangle .
thus for samples with scharp corners the superconducting / normal transition moves to higher field ( for fixed surface area ) @xcite .
multivortex states can nucleate in the disk for vorticity @xmath52 and @xmath53 and in the square and the triangle for @xmath54 and @xmath55 .
moreover , for the disk , with increasing field , the multivortex state always transits to a giant vortex state for fixed @xmath3 , while for the square and the triangle some @xmath3-states are multivortex states over the whole magnetic field range . for the triangle
this is the case for the states with vorticity @xmath56 and for the square for the states with @xmath57 and @xmath53 .
this indicates that breaking the axial symmetry favors the multivortex state over the giant vortex state . in some magnetic field regions ,
the vortex states exhibitis a paramagnetic response , i.e. @xmath58 .
this occurs in the disk for metastable states with @xmath59 , @xmath60 and in the square for metastable states with @xmath61 , @xmath62 , @xmath53 . for the triangle @xmath63
is always positive , i.e. only diamagnetic behaviour is observed . in fig .
[ flux](a ) we plot the magnetic flux @xmath64 passing through the superconductor at the thermodynamic transition field @xmath65 as a function of the vorticity @xmath3 for the three geometries .
we scaled this flux by @xmath66 which is the expected flux when no boundary effects are important .
the result for the disk is given by circles , for the square by squares and for the triangle by triangles .
in the figure we use open symbols for transitions between two giant vortex states , filled symbols for transitions between a multivortex and a giant vortex state , and crossed symbols for transitions between two multivortices . the symbol for
the largest @xmath3 corresponds to the superconducting / normal transition . for the disk we find @xmath67 transitions between a giant vortex and a multivortex state for @xmath61 and @xmath68 , while the other transitions occur between two giant vortex states . for the square
we find transitions between a multivortex state and a giant vortex state for @xmath69 , @xmath70 , @xmath55 and between two multivortex states for @xmath57 and @xmath71 .
for the triangle , transitions between a giant and a multivortex state occur for @xmath61 , @xmath68 and @xmath55 and between two multivortex states for @xmath69 , @xmath72 and @xmath62 .
notice further that for @xmath73 , the value of @xmath74 becomes almost independent of @xmath3 for the three geometries : @xmath75 for the disk and the square and @xmath76 for the triangle .
the magnetic flux @xmath64 passing through the superconductor at the thermodynamic transition field @xmath65 can be fitted for the three geometries by the following formula @xmath77 with @xmath78 @xmath79 and @xmath80 for the disk ( solid curve in fig .
[ flux](a ) ) , @xmath81 @xmath82 and @xmath83 for the square ( dashed curve ) , and @xmath84 @xmath85 and @xmath86 for the triangle ( dotted curve ) . the amount of flux increase needed to increase the vorticity of the ground state from @xmath3 to @xmath87 is plotted in fig .
[ flux](b ) for the disk ( solid curves , open circles ) , the square ( dashed curve , open squares ) and the triangle ( dotted curve , open triangles ) . for @xmath88 , this value is almost independent of @xmath3 and of the geometry and equals @xmath89 , i.e. it is approximately equal to , but still different from , the flux quantum @xmath90 .
notice that @xmath91 for any @xmath3 which is a consequence of the demagnetization effects . to distinguish whether the superconducting state is a multivortex state or a giant vortex state and to determine the multivortex to giant vortex state transition field , we considered the cooper - pair density @xmath92 in the center of the superconductor @xcite
. we can be sure that the superconducting state is a multivortex state if @xmath93 for @xmath94 .
the reason is that giant vortices are always in the center of the superconductor , and hence @xmath95 for giant vortex states .
[ denscent ] shows the cooper - pair density in the center of the square as a function of the applied magnetic field .
the cooper - pair density @xmath92 is finite for @xmath96 and @xmath97 for @xmath69 and @xmath70 , respectively , and @xmath98 for @xmath99 . for @xmath57
the cooper - pair density in the center differs from zero over the whole magnetic field region where the @xmath57 state is stable . on the other hand @xmath98
does not guarantee that the superconducting state is a giant vortex state .
for example , the multivortex state with @xmath100 in a square shows 4 vortices away from the center situated on the diagonals and one vortex in the center , and hence @xmath101 .
therefore , we studied the cooper - pair density distribution in detail .
if two vortices are very close to each other , then the cooper - pair density on the axis between these two vortices will become very low too , which means that the separation between two vortices becomes invisible in the contourplots .
therefore we have to define another criterion to determine the multivortex to giant vortex transition .
if the maximum between two minima in the cooper - pair density ( i.e. the vortices ) is lower than @xmath102 of the maximum cooper - pair density @xmath103 in the sample , then we will say that the vortices form a giant vortex state instead of a multivortex state . with this criterion
we find that the @xmath100 state in a square is always a multivortex state and the @xmath104 state is a multivortex state for @xmath105 and a giant vortex state for @xmath106 .
how do the multivortex states look like ? figs .
[ contv](a - c ) show the cooper - pair density for a multivortex state with @xmath69 , @xmath70 , and @xmath62 at @xmath107 , @xmath108 , and @xmath109 respectively .
high cooper - pair density is given by dark regions , low cooper - pair density by light regions .
for @xmath69 the vortices are along the diagonal , for @xmath110 the vortices are on a triangle , and for @xmath57 they are on a square . for @xmath57
only the multivortex state is found which is favoured over the giant vortex state .
the reason is that the square vortex lattice easily fits in the sample
[ contv](d , e ) show the phase of the order parameter for the multivortex states with @xmath100 at @xmath111 and with @xmath104 at @xmath112 .
phases near zero are given by light regions and phases near @xmath113 by dark regions . by going around the superconductor , the phase cleary changes 5 times @xmath113 in fig .
[ contv](d ) and 6 times @xmath113 in fig .
[ contv](e ) . for @xmath100
there are 4 vortices on a square and the fifth vortex is in the center .
the latter has clearly vorticity one . for @xmath114
there are also 4 vortices on a square and the other vortices are in the center forming one giant vortex with vorticity 2 .
thus in this case we have the remarkable coexistence of a giant vortex in the center with vorticity 2 and 4 clearly separated vortices around it . for this case the multivortex to giant vortex transition field
is defined as the field where separate vortices appear with decreasing field .
this means that in fig .
[ emag ] some states are indicated as multivortex , even though there exist a ( giant ) vortex in the center with vorticity @xmath94 .
thus we consider a state no longer as a giant vortex state when not all flux of the vortex is confined in a single connected region .
notice also that not only the configuration of the multivortices tries to have the same geometry as the sample , but also the giant vortex geometry depends on the sample geometry .
we found that for this size of the square sample , the states with @xmath73 are always in the giant vortex state . with increasing @xmath3
this giant vortex grows and superconductivity only occurs in the corners of the square .
this is illustrated in figs .
[ contv11](a , b ) which shows the cooper - pair density for the @xmath48 state at @xmath115 and @xmath116 .
it is obvious that further increasing the field pushes the superconducting condensate more to the corners . at the superconducting / normal transition field
@xmath117 the corners becomes normal too .
[ squarel3 ] shows the positions of the vortices for the @xmath110 state in a square at applied magnetic fields @xmath118 , @xmath119 , @xmath120 , and @xmath121 . the latter one is just below the multivortex to giant vortex state transition field @xmath122 .
the solid lines indicate the square boundaries .
with increasing field the vortices move towards the center of the square and at @xmath122 they combine in the center and form one giant vortex with vorticity @xmath123 in the multivortex state , one vortex is always situated on the diagonal of the square , regardless of the magnetic field .
the other two vortices are located such that the three vortices form a equilateral triangle which is centered in the center of the square .
since the vortices move to the center with increasing field , the width @xmath124 of the triangular vortex lattice decreases , i.e. @xmath125 @xmath126 , @xmath127 , @xmath128 at @xmath118 , @xmath119 , @xmath129 , @xmath121 , respectively . for the triangle geometry multivortex states
nucleate with vorticity @xmath130 and @xmath55 ( see figs .
[ emag](e , f ) ) . figs .
[ contd](a - c ) show the cooper - pair density for a multivortex state with @xmath69 , @xmath70 , and @xmath62 at @xmath131 , @xmath132 , and @xmath109 respectively .
high cooper - pair density is given by dark regions and low cooper - pair density by light regions .
in the multivortex state with vorticity @xmath69 the vortices are situated along one of the perpendicular bisectors of the triangle . in the @xmath133 state the vortices are on a triangle which easily fits in the sample ,
while the @xmath57 state consists of 3 vortices on a triangle and the fourth vortex is situated in the center . instead of the square configuration as in the case of the square geometry , the vortex lattice tries to copy the geometry of the sample , i.e. the triangular geometry . for the multivortex states with @xmath100 and @xmath104 ,
the separation of vortices becomes invisible in the contourplots of the cooper - pair density , which show one big vortex in the center .
the reason is that the maximum cooper - pair density on the axis between two vortices is very low .
we show the phase of the order parameter in figs .
[ contd](d , e ) for the multivortex states with @xmath100 at @xmath134 and with @xmath104 at @xmath135 .
phases near zero are given by light regions and phases near @xmath113 by dark regions .
in both cases there is a coexistence of a giant vortex in the center and 3 separated multivortices around it placed in the direction of the corners . taking a loop around the giant vortex , the phase changes respectively by 2 and 3 times @xmath113 , which means that the vorticity of the giant vortex is 2 in the case of the @xmath100 multivortex state and 3 in the case of the @xmath104 multivortex state .
notice also that for @xmath104 the geometry of the giant vortex is not axial symmetrical , but triangular .
states with @xmath73 are always giant vortex states as we also found for the square . with increasing @xmath26
this giant vortex grows and for large vorticities superconductivity only occurs in the corners of the triangle . further increasing
the field pushes the superconductivity more to the corner untill these corners becomes normal too at the superconducting / normal transition field . for the circular disk
we find multivortex states with vorticity @xmath52 and @xmath53 ( see figs .
[ emag](a , b ) ) . figs .
[ contc](a - d ) show the cooper - pair density for the multivortex states with vorticity @xmath54 at @xmath136 , @xmath119 , @xmath137 and @xmath132 , respectively .
high cooper - pair density is given by dark regions , low by light regions .
multivortex states for disks were already studied in previous papers ( see introduction ) .
therefore , in this paper we only stress that in a disk the multivortices are positioned on a ring , which means that also in this case the sample imposes its symmetry on the vortex lattice . from the study of the cooper - pair density and the phase of the order parameter we learned that : ( i ) multivortex states nucleate in disks as well as in squares and triangles for several values of the vorticity @xmath3 , and ( ii ) the vortex lattices tries to have the same geometry as the sample .
since we studied samples with finite thicknesses , demagnetization effects are important and therefore we had to solve for the magnetic field distribution around the sample . in this section
we will describe the magnetic field distribution for the square .
the results for the disk and the triangle are analogous . in figs .
[ magv23](a - f ) the magnetic field distribution is shown for the square geometry for the state with vorticity @xmath69 at @xmath138 , @xmath139 and @xmath119 ( see open circles in fig . [ denscent ] ) , and with vorticity @xmath110 at @xmath140 , @xmath141 and @xmath132 , respectively .
high magnetic field is given by dark regions and low by light regions .
the magnetic field is clearly non uniform in and around the sample .
the dark spots in the square are the vortices and the dark regions near the sample surface are due to the compression of the magnetic lines when they are forced to go around the sample .
these regions are responsible for the demagnetization effects .
it is clear that with increasing external field and fixed number of vortices , the demagnetization effects are more pronounced , because the superconductor has to expel more magnetic field . in figs .
[ magv23](a , b , d , e ) the superconducting state is a multivortex state and the separated vortices are clearly visible , while in figs .
[ magv23](c , f ) where @xmath142 and @xmath143 for @xmath69 and @xmath144 , respectively , there is one giant vortex in the center . with increasing field the vortices move towards the center and at @xmath145 they combine to one giant vortex state .
notice that the giant vortex state is not necessarily axial symmetric as was in the case of the disk . in figs .
[ magv4](a - d ) the magnetic field distribution is shown for the square geometry for the state with vorticity @xmath57 at the magnetic fields indicated by the open circles in fig .
[ denscent ] , i.e. @xmath146 , @xmath132 , @xmath147 and @xmath148 , respectively .
now , there is no transition from multivortex to giant vortex state and the four vortices are cleary visible as the dark spots .
notice that from the magnetic field distribution one clearly observes that the vortex lattice is a square lattice , i.e. the lattice geometry is the same as the sample geometry , and that the vortices move towards the center with increasing field . for the multivortex states with higher vorticity ,
the separated multivortices are not visible anymore in the contour plots of the magnetic field distribution .
the problem is the same as for the contour plots of the cooper - pair density , i.e. the vortices are too close to each other and the spots corresponding to high magnetic fields are overlapping in the picture . for high vorticity and high external fields
, the total magnetic field appreciably differs from the externally applied field only in the corners of the square .
[ magv11](a , b ) shows the magnetic field distribution for the same configuration as in figs .
[ contv11](a , b ) , i.e. the @xmath48 state at @xmath115 and @xmath116 , respectively . a local decrease in magnetic field
is given by the light regions , an increase by the dark regions .
in both pictures the magnetic field is only substantially expelled in the corners and consequently only near the corners there is a higher density of magnetic field lines at the outside of the square . further increasing the field
destroys the superconductivity and , thus , the total field becomes equal to the external one over the whole sample .
next , we investigate the dependence of the magnetic field on @xmath19 . figs .
[ magneetz](a - f ) show the magnetic field distribution for the @xmath57 state in a square for different values of @xmath149 @xmath150 , @xmath151 , @xmath152 , @xmath153 , @xmath154 and @xmath155 , respectively .
the applied magnetic field is @xmath156 .
high magnetic field is given by dark regions , low magnetic field by light regions .
in the plane of the superconductor , i.e. @xmath157 , the magnetic field which penetrates the superconductor is either compressed into multivortices or expelled to the outside of the sample . therefore , the four dark spots in fig . [ magneetz](a ) indicate the vortices and the light regions towards the sample boundary are due to the expulsion of the magnetic field towards the outside of the superconductor . as a consequence , the magnetic field increases in a small strip near the sample boundary . with increasing @xmath19 and @xmath158 , the magnetic field will still be influenced by the superconductor .
the demagnetization effects decrease with increasing @xmath19 and the compression of the magnetic field lines into vortices becomes smaller .
therefore , the vortices and the expulsion of the field will become less pronounced with increasing @xmath19 . at @xmath159 ,
the vortices and the results of the magnetic field expulsion are still visible by the dark and light regions ( see fig .
[ magneetz](b ) ) . in fig .
[ magneetz](c ) and ( d ) , at @xmath160 and @xmath161 , respectively , the contrast in the picture decreases which means that the influence of the superconductor , i.e. the compression and expulsion of the magnetic field lines , decreases . at @xmath162 the magnetic field
just slightly decreases right above the superconductor compared to the external field ( see fig . [
magneetz](e ) ) . at @xmath163
the total magnetic field is homogeneous . far away from the superconductor ,
the magnetic field is not influenced by the superconductor and equals the external field .
this is clearly shown in fig .
[ magneetz](f ) .
when a superconducting sample is placed in an external magnetic field , the magnetic field is expelled from the superconductor due to screening currents near the sample boundary .
the direction of the screening currents is such that the corresponding magnetic field is opposite to the external one , which leads to a lower total field in the superconductor .
magnetic field penetrating the superconductor creates currents flowing in the other direction than the screening currents .
the competition between these currents and the screening currents results in the existence of vortices .
[ currv](a - d ) show vector plots of the supercurrent in the superconducting square for the @xmath61 state at @xmath164 , the @xmath69 state at @xmath138 , the @xmath110 state at @xmath165 and the @xmath57 state at @xmath166 , respectively . figs .
[ currv](e - h ) show the corresponding contour plots of the phase of the order parameter .
phases near @xmath113 are given by dark regions , phases near zero by light regions . from the phase of the order parameter one
can easily determine the number and the positions of the vortices .
[ currv](a ) it is clear that the screening currents near the sample boundary flow clockwise and the currents around the vortex in the center counterclockwise . in figs .
[ currv](b , c , d ) there are currents flowing counterclockwise around @xmath167 @xmath70 and @xmath62 vortices , respectively . around one vortex , the size of the current , indicated by the length of the arrows in figs .
[ currv](a - d ) , is not the same for every angle . in fig .
[ currv](b ) it is clear that in the region between the two vortices the currents around these two vortices cancel each other out .
also , in the case of @xmath110 and @xmath57 the currents around the different vortices cancel out each other in the center of the sample ( see fig . [ currv](c , d ) ) . from figs .
[ currv](a - d ) one expects anti - vortices towards the corners , because there are some spots where the currents flow in clockwise direction . that these are not really anti - vortices can be seen from the phase of the order parameter ( figs . [ currv](e - h ) ) . by going around an anti - vortex ,
the phase changes with @xmath168 and this is clearly not the case here .
moreover , the cooper - pair density , shown in figs .
[ contv](a - c ) , is not zero at these positions .
next , we investigate the superconducting current density in the triangular sample . figs .
[ currd](a - c ) show vector plots of the supercurrent in the superconducting triangle for the @xmath61 state at @xmath164 , the @xmath169 state at @xmath136 , and the @xmath110 state at @xmath170 , respectively . figs .
[ currd](d - f ) show the contour plots of the corresponding phase of the order parameter .
phases near @xmath113 are given by dark regions , phases near zero by light regions .
the behavior of the supercurrent in triangular samples is similar to the one in square samples .
the screening currents flow clockwise and the current around the vortices in the opposite direction .
the currents around different vortices cancel each other in the region between them . towards the corners ,
there are some spots where the current flows in a clockwise direction , but these spots are not anti - vortices .
this can be seen from the phase of the order parameter ( see figs . [ currd](d - f ) ) and from the cooper - pair density ( see figs . [ contd](a - b ) ) .
not only the stability region of the multivortex states with respect to the giant vortex states depends on the sample geometry , but also the stability of each individual superconducting state is sensitive to the geometry . in fig .
[ stab ] we show the magnetic field range @xmath171 over which the vortex state with vorticity @xmath3 is stable , i.e. @xmath172 ( see also figs .
[ pen_exp](a , b ) ) , as a function of the vorticity @xmath3 , for @xmath173 and in the inset for @xmath174 . for the disk
the result is shown by the open circles , for the square by the open squares and for the triangle by the open triangles where the curves are guides to the eye .
for the circular disk the stability region @xmath175 uniformly decreases with increasing @xmath3 with a slight dip at @xmath176 .
the square and the triangle exhibit a peak structure in the region @xmath177 .
for the square we find that the state with @xmath57 is stable over a larger magnetic field region than the state with vorticity @xmath110 , which is a consequence of the fact that the vortex lattice tries to keep the same geometry as the sample .
for the triangle we find a peak at @xmath110 and a dip at @xmath69 for the same reason .
notice that : i ) the peak structure is more pronounced for structures which fit more closely the triangular abrikosov lattice ; ii ) for @xmath178 no clear peaks are found ; iii ) the vortex states in the square and circle geometry have almost the same stability range for @xmath179 and @xmath174 ; iv ) for @xmath180 the stability range for the vortex state in the triangular geometry becomes substantially smaller than for the other two geometries which have less sharp corners .
thus sharp corners decrease the stability range of the vortex states , i.e. it is easier for flux to enter the system ; and v ) fig .
[ flux ] , which shows the extra flux needed to increase the vorticity by one unit , contains complementary information to fig .
[ stab ] .
untill now , all our calculations were done for fixed temperature @xmath181 .
now we will include temperature and our lateral dimensions and fields will be expressed in the zero temperature results @xmath182 and @xmath183 , respectively .
temperature will be expressed in units of the zero magnetic field criticial temperature @xmath39 .
we take the surface area of our samples @xmath184 and the thickness @xmath185 .
the @xmath0 phase diagram is shown in the inset of fig .
[ htfase ] for the disk ( solid curves ) , the square ( dashed curves ) and the triangle ( dash - dotted curves ) for the states with vorticity @xmath186 and @xmath61 , thus for low fields and temperatures close to @xmath39 .
the thick curves are the superconducting / normal transitions and the thinner curves indicate the expulsion and the penetration fields , i.e. the boundaries of the stabilitiy region of the state with vorticity @xmath3 .
the lower thin curves show the transition from the state with vorticity @xmath61 to @xmath186 with decreasing field ( expulsion ) and the upper thin curves show the transition from @xmath186 to @xmath61 with increasing field ( penetration ) .
[ htfase ] shows the @xmath187 phase diagram for higher fields . for the sake of clarity
only the superconducting / normal transition @xmath188 is shown as a function of temperature .
the black dots indicate the transitions fields between the different @xmath3-states . for every ( fixed ) temperature
the superconducting / normal transition field is highest for the triangle and lowest for the disk .
for every ( fixed ) magnetic field , the critical temperature is highest for the triangle and lowest for the disk .
this means that with sharper corners , the critical temperature and critical field are enhanced due to an enhanced surface superconductivity .
these results are in good agreement with the phase diagrams found in ref . @xcite .
we investigated theoretically the influences of the geometry of thin superconducting samples on the vortex configuration .
therefore , we considered superconducting disks , squares and triangles with the same surface area @xmath41 and the same thickness @xmath42 for @xmath189 .
for these three geometries we calculated the free energy and the magnetization of the different giant and multivortex states as a function of the applied magnetic field and we indicated the multivortex to giant vortex transitions for fixed vorticity @xmath3 .
multivortex states were found for disks as well as for squares and triangles for several values of the vorticity . for given @xmath3 ,
the vortex lattice was different in the three geometries due to the fact that it tries to adapt to the geometry of the sample .
this influences considerably the stability range of the different vortex states . for squares and triangles we found magnetic field regions where there is a coexistence between a giant vortex state in the center and several separated vortices in the direction of the sample corners . near the superconducting / normal transition
we do not find multivortices , anti - vortices or a combination of them , but we find surface superconductivity .
we studied the magnetic field distribution across the superconductor and around the superconductor which clearly shows the demagnetization effects , which are very important for samples of finite thickness .
the vector plots of the superconducting current showed spots where the current flows in clockwise direction . from the phase of the order parameter and the cooper - pair density
we conclude that these spots are not anti - vortices .
we also investigated the stability of the vortex states with vorticity @xmath3 by calculating the magnetic field range over which the vortex states with vorticity @xmath3 are stable .
we found that this stability range sensitively depends on the sample geometry . as a function of @xmath3 we found enhanced stability for the triangle for @xmath110 and for the square for @xmath57 . in the last section ,
we also included temperature by calculating a @xmath0 phase diagram for the disk , the square and the triangle . with sharper sample corners , we found that for fixed temperature , the superconducting / normal transition field @xmath188 moves to higher fields , and for fixed field , the critical temperature increases .
this work was supported by the flemish science foundation ( fwo - vl ) , the onderzoeksraad van de universiteit antwerpen ( goa ) , the interuniversity poles of attraction program - belgian state , prime minister s office - federal office for scientific , technical and cultural affairs , and the european esf - vortex matter .
discussions with s. yampolskii are gratefully acknowledged . | the influence of the geometry of a thin superconducting sample on the penetration of the magnetic field lines and the arrangement of vortices are investigated theoretically .
we compare superconducting disks , squares and triangles with the same surface area having nonzero thickness .
the coupled nonlinear ginzburg - landau equations are solved self - consistently and the important demagnetization effects are taken into account .
we calculate and compare quantities like the free energy , the magnetization , the cooper - pair density , the magnetic field distribution and the superconducting current density for the three geometries . for given vorticity
the vortex lattice is different for the three geometries , i.e. it tries to adapt to the geometry of the sample .
this also influences the stability range of the different vortex states . for certain magnetic field
ranges we found a coexistence of a giant vortex placed in the center and single vortices toward the corners of the sample . also the @xmath0 phase diagram is obtained .
[ theorem]acknowledgement [ theorem]algorithm [ theorem]axiom [ theorem]claim [ theorem]conclusion [ theorem]condition [ theorem]conjecture [ theorem]corollary [ theorem]criterion [ theorem]definition [ theorem]example [ theorem]exercise [ theorem]lemma [ theorem]notation [ theorem]problem [ theorem]proposition [ theorem]remark [ theorem]solution [ theorem]summary |
[ sec1 ] consider the following problem : @xmath0 random walkers set off from the vicinity of the origin , in @xmath11-dimensional euclidean space , at time @xmath12 .
they are divided into @xmath2 different families : the number of walkers in the @xmath13th family is @xmath14 , so that @xmath15 . within a particular family ,
walkers are indifferent to each other : their paths may cross .
however , each family behaves viciously towards all the others : if two walkers from different families meet , both are annihilated .
we may ask many different questions about this problem , but a fundamental quantity is the probability @xmath16 that all the walkers have still survived up to time @xmath3 .
equivalently , we may consider the ensemble of @xmath0 independent random walks : @xmath16 is the fraction of these in which no walkers of different families have intersected up to time @xmath3 . for a discrete time process on a lattice , if @xmath17 is the position at time @xmath3 of the @xmath18th walker of the @xmath13th family , then @xmath16 is the expected value of the indicator function @xmath19 this problem is of interest for several reasons .
it generalises a number of cases : 1 .
@xmath20 ( @xmath21 ) corresponds to _ vicious walkers _ _ , a term first introduced by fisher @xcite .
it has been studied using @xmath22-expansion methods @xcite similar to those of the present paper .
the survival probability is known exactly for @xmath23 in the case when all walkers have the same diffusion constants : it decays as a power @xmath24 @xcite .
these methods rely essentially on a fermionic description of the problem @xcite .
krattenthaler et al.@xcite introduced the method of the schur functions and katori and tanemura @xcite developed this and discussed the relation with the random matrix theory.these methods do not appear to extend to the case when the diffusion constants are different .
results in this case have been reported for @xmath25 @xcite . _ _ _ 2 .
the case @xmath25 , with @xmath26 and @xmath27 , has been studied by krapivsky and redner@xcite as a model of @xmath28 predators ( ` lions ' ) hunting a prey ( ` lamb ' ) .
they were able to obtain exact results for the asymptotic behaviour of the survival probability , again in @xmath23 , for the cases @xmath29 and arbitrary diffusion constants .
for general @xmath28 , the exponent is related to the smallest eigenvalue of a dirichlet problem in a certain @xmath30-dimensional compact region , and is probably not expressible analytically , but for large @xmath28 these authors were able to estimate its behaviour .
the ` lion - lamb ' problem for @xmath23 is related to a version of the ` ballot problem ' in which it is required to know the probability that one candidate in a ballot remains ahead of the @xmath28 others at all stages in the poll .
exact results are known only for @xmath31 @xcite .
the ` lion - lamb ' problem has another interpretation , in terms of _ multiscaling _ _ : if we first regard the trajectory @xmath32 of the lamb as fixed , and if @xmath33 is the probability that it has not been met by a _
lion , then @xmath34 where the average is over all the realisations of @xmath32 .
the fact that @xmath35 decays with @xmath3 with an exponent which is not simply linear in @xmath28 is symptomatic of multiscaling in this problem . _
_ _ _ _ 4 .
more generally , we can regard @xmath36 as being the average of the @xmath37th power of the survival probability of a _ single _ walker of family 1 , in the presence of @xmath38 walkers of the @xmath39 other families .
our problem has a strong resemblance to that of the intersection probabilities of brownian paths . in this case , one studies the ensemble of @xmath0 random walks in @xmath11 dimensions each of which begin a distance @xmath40 from the origin and which arrive on a hypersphere of radius @xmath41 before they cross @xmath42 , _ irrespective _ of how long this takes .
once again the walkers are divided into families , and in this case one is interested in the probability @xmath43 that the _ paths _ of walkers of different families do not intersect .
thus , instead of ( [ indic ] ) , @xmath44 is the expected value of @xmath45 and it is supposed to decay as @xmath46 as @xmath47 , where @xmath48 depends nontrivially on the @xmath49 .
this problem is trivial in @xmath23 , and turns out to have an upper critical dimension @xmath50 , below which an @xmath22-expansion is possible@xcite . for @xmath51
an exact formula for @xmath52 has been derived@xcite , by exploiting the conformal invariance of the problem . _ _ _ _ given these remarks , it seems important to investigate the general case described in the opening paragraph . as far as we know ,
the fermionic methods used to attack the vicious walker problem for @xmath23 do not extend to this case . we have therefore employed a renormalisation group ( rg ) method , which yields , for @xmath5 , results for the exponent @xmath53 of the power law decay of @xmath16 as a power series in @xmath54 . by using field - theoretic methods , the calculation is streamlined , and , once the formalism is set up , involves relatively little explicit calculation .
we have carried this computation through @xmath7 , and for arbitrary diffusion constants of each family .
it would be tedious , but not difficult , to carry it further , as the actual feynman integrals are elementary .
we also show that in two dimensions @xmath16 decays as a universal power of @xmath55 .
the layout of this paper is as follows : in sec .
[ sec2 ] , for completeness , we collect all our results and show how they reduce in the above - mentioned special cases . in sec .
[ sec3 ] we set up the field - theoretic formulation of the problem , then in the next section carry out the rg analysis . sec .
[ sec5 ] contains a summary and further remarks .
several of the detailed calculations are relegated to appendices .
let @xmath2 be the number of families , @xmath14 be the number of walkers in the @xmath13th family , and @xmath56 be their diffusivity .
let @xmath16 be the survival probability @xmath57\ ] ] in this case there is a finite probability that any pair of walkers will never meet . as a result ,
@xmath16 approaches a non - universal constant value less than 1 , with leading power - law corrections of the form @xmath58 .
@xmath59 where , with @xmath8 , @xmath60 with @xmath61 where @xmath62 and @xmath63 from this may be deduced various special cases : assume that @xmath64 for all @xmath65 . then @xmath66 with @xmath67
note that these are expressed in terms of symmetric polynomials in the @xmath49 .
this in fact holds to all orders in @xmath22 . when @xmath68 @xmath69 should be equal to the _ survival exponent _ @xmath70 of the vicious walkers . in this case
@xmath71 for @xmath72 and the result ( [ eqn : result1 ] ) gives @xmath73 the result ( [ eqn : result1b ] ) gives @xmath74 this agrees with the result reported as eqn .
( 5.2 ) , with eqn . ( 3.13 ) in mukherji and bhattacharjee@xcite ( see also @xcite ) .
it has been proved that @xcite @xmath75 note that although this exact result agrees with that from the first - order @xmath22-expansion ( [ eqn : result2b ] ) on setting @xmath76 , this is probably fortuitous , as , in the case of unequal diffusivities the exact result depends on their ratio , while the first - order term in ( [ eqn : result2 ] ) does not .
the ` @xmath28 lions and one lamb ' problem studied by krapivsky and redner @xcite is a special case of the present model in which @xmath77 in this case @xmath78 for @xmath72 and the result ( [ eqn : result1 ] ) gives @xmath79 where @xmath80 .
redner and krapivsky @xcite reported the exact solution for @xmath81 in @xmath23 ( i.e. , @xmath76 ) , @xmath82^{-1}. \label{eqn : rk1}\ ] ] it was shown that @xmath83 is monotonically decreasing in @xmath84 and @xmath85 if we neglect @xmath86 and set @xmath87 in ( [ eqn : result3 ] ) , we have @xmath88 which is monotonically decreasing in @xmath84 and @xmath89 in this case , there is a logarithmic decay with universal exponent : @xmath90 where @xmath91 note that this is independent of the @xmath56 ( as long as no pair of them both vanish ) : the dependence shows up only in the prefactor and the non - leading terms .
in this section , we set up the general problem as a continuum field theory , so that the powerful techniques of the field - theoretic rg may be applied .
the general method for formulating such stochastic particle systems as field theories , as originally proposed by doi@xcite and peliti@xcite , has been described at length elsewhere@xcite and we shall only summarise how this is applied in the case of interest . initially , the problem is formulated on a lattice , for example @xmath92 , the sites of which are labelled by a vector @xmath93 .
the microstate of the system at a given time is specified by occupation numbers @xmath94 , which specify that there are @xmath95 walkers of family @xmath13 at the site @xmath93 .
note that we treat walkers of the same family as identical particles : this makes no difference in the problem of interest .
to each microstate is associated a vector in a fock space @xmath96 , built by applying raising operators to the vacuum , or empty state , @xmath97 : @xmath98 where @xmath99=\delta_{jk}\delta_{{\bf r } { \bf r}'}$ ] , and @xmath100 .
let @xmath101 be the probability of finding the system in this microstate at time @xmath3 , and define the state @xmath102
latexmath:[\[|\psi(t)\rangle\equiv\sum_{\{m_j({\bf r})\}}p(\{m_j({\bf r})\};t ) then the master equation , which is linear equation describing the time - evolution of the probabilities @xmath101 , is equivalent to the schrdinger - like equation @xmath104 where @xmath105 may be expressed explicitly in terms of the raising and lowering operators . for the case of independent random walks in continuous time ,
@xmath106 where @xmath107 is the lattice spacing , and the sum is over nearest neighbour pairs of sites @xmath108 .
the probability of finding the walkers at sites @xmath109 at time @xmath3 ( where @xmath110 with @xmath21 ) is then given by @xmath111 of course , when this is summed over all the @xmath109 , it gives unity . before considering how to implement the non - intersection constraint ,
let us first discuss the continuum limit and the path integral representation . in this non - interacting case
, the continuum limit may be taken rigorously .
the raising and lowering operators go over into ( distribution - valued ) field operators satisfying @xmath112=\delta_{jk}\delta({\bf r}-{\bf r}')$ ] , and the generator of time evolution becomes @xmath113d^d\!r\ ] ] since the walkers are all supposed to begin in the vicinity of the origin at @xmath12 , that is , a finite number of lattice spacings away , in the continuum limit @xmath114 @xmath115 the path integral representation is derived@xcite by breaking the time interval @xmath116 into slices of length @xmath117 , so that the time - evolution operator @xmath118 is the product of factors @xmath119 , and inserting a complete set of coherent states at each time slice
. this has the effect of replacing the operators @xmath120 and @xmath121 by time - dependent @xmath122-number fields @xmath123 and @xmath124 respectively .
after taking the limit @xmath125 , the matrix element ( [ prob ] ) becomes a functional integral @xmath126 where @xmath127dtd^d\!r\ ] ] and @xmath128 .
we now discuss how to incorporate the constraint that walkers of different families should not meet . rather than insert the indicator function ( [ indic ] ) into the path integral , it more convenient to consider a slightly more general problem in which , before taking the limits @xmath114 and @xmath125 , each set of trajectories
is weighted by a factor @xmath129 where the @xmath130 are a set of dimensionless parameters ( the factors of @xmath107 and @xmath117 are inserted to make the continuum limit simpler . ) the case of strict non - intersection corresponds to the limit @xmath131 .
however , we shall show that , for @xmath132 , the leading behaviour is independent of the precise value of these parameters ( as long as they are all strictly positive ) and , moreover , the rg fixed point , at which non - leading corrections to the asymptotic behaviour disappear , corresponds to the limit of infinite @xmath133 . in the formal continuum limit , this corresponds to a modification of the action in the path integral @xmath134 the feynman rules for this theory are very simple@xcite and are illustrated in fig .
[ figrules ] .
we denote averages and correlations with respect to the bare action @xmath135 by the subscript @xmath136 : averages with respect to the full action @xmath137 are denoted by @xmath138 . * the fourier - laplace transform of the bare propagator @xmath139 is represented by a line directed towards increasing time ( conventionally , right - to - left ) . in the @xmath140 representation
the bare propagator is simply @xmath141 ; * the interaction @xmath142 is represented by a vertex with one incoming and one outgoing pair of lines each of type @xmath143 and @xmath144 ; * as usual , wave number @xmath145 and ( imaginary ) frequency @xmath146 are conserved at the vertices , and internal loop integrations @xmath147 and @xmath148 are carried out . the survival probability is now given by the correlation function @xmath149 evaluated with the weight @xmath150 .
however , this does not exist in the formal continuum limit , because the perturbative feynman diagram expansion of @xmath151 contains ultraviolet ( short - distance or short - time ) divergences .
physically this is because two walkers , having interacted once , are , in the continuum limit , likely to interact an infinite number of times as @xmath125 .
this divergence may be regulated , either by imposing an explicit cut - off @xmath152 in the feynman integrals , or , more easily , by dimensional regulation . for @xmath132
this field theory is renormalisable : the singular dependence on the regulator may be absorbed into a finite number of parameters . in the case of the theory of interest
, this procedure is particularly simple@xcite : no renormalisation of the field @xmath153 nor of the diffusion constants is required , only a simple renormalisation of the coupling constants @xmath154 , which can be computed exactly to all orders .
the lack of field and diffusion constant renormalisation holds mathematically because there are no loop corrections to the propagators .
physically it is because an isolated walker does not interact , even with itself , in the absence of any branching processes .
when the coupling constant renormalisation is done , all correlation functions of products of @xmath155 and @xmath156 , at _ distinct _ space - time points , have a finite limit as the regulator is removed , when expressed in terms of the renormalised couplings . the fact that the renormalised theory must be defined at some arbitrary scale then leads to rg equations . _ _ however , this procedure is not sufficient to render finite correlation functions involving so - called _ composite operators _ like @xmath157 .
physically , this is because if the walkers all begin at exactly the same point , they will all annihilate each other immediately ! in order to obtain finite renormalised correlation functions , it is first necessary to point - split the fields : @xmath158 ( note that it is not necessary to split the starting points of walkers of the same family , since they do not interact . )
now consider a correlation function of this product with an arbitrary product @xmath159 of fields whose time arguments are all strictly positive : @xmath160 in the cut - off theory , we could simply make a taylor expansion of this in powers of the @xmath161 .
this would have the form @xmath162 where the summation is taken over all possible derivatives @xmath163 of @xmath164 with respect to the @xmath165 . on the basis of dimensional analysis , the first term gives the leading behaviour for @xmath166 as @xmath167 , at least at @xmath51 . in the interacting theory , however , each term in ( [ taylor ] ) has to be renormalised separately . as a result @xmath168 where all the correlation functions are finite as the regulator is removed
since each of the @xmath169 may acquire a nontrivial anomalous dimension through this procedure , the renormalised functions @xmath170 have a non - trivial dependence on their arguments .
the important feature of ( [ ope0 ] ) is that the renormalisation constants @xmath171 are independent of @xmath159 .
this we may write the _ operator product expansion _ ( ope ) @xmath172 where the ope functions @xmath173 , _ etc . _ are in general nontrivial . _ _ _ _ _ _ each term in the ope ( [ ope ] ) , when substituted into @xmath151 , will give rise to nontrivial power - law dependence on @xmath3 for @xmath132 .
however , since in the noninteracting theory we know that only the first term is important as @xmath167 , we shall assume that this remains true for sufficiently small @xmath22
. further discussion of this point will be postponed to sec .
[ sec5 ] . in the absence of any interactions , the survival probability @xmath174 , as can be seen by evaluating the first diagram in the expansion shown in fig .
[ figgbar ] with all the wave numbers @xmath175 set to @xmath176 .
for @xmath4 , the higher order terms give contributions which correct this constant , as well as terms which are subleading as @xmath167 .
for example , a typical one - loop diagram like that in fig .
[ figgbar ] is proportional to @xmath177 this integral diverges at large @xmath178 for @xmath179 , a consequence of taking the naive continuum limit .
if we impose a cut - off @xmath180 , the leading term behaves as @xmath181 , with a non - universal coefficient , while the remainder is finite as @xmath182 and behaves as @xmath183 .
the non - universal constant term corresponds to a finite probability that the walkers annihilate at short times , before escaping each other .
this behaviour persists at higher orders in the interactions . for @xmath132
, however , each successive term in the bare perturbation grows as a larger and larger positive power of @xmath3 , and it is necessary to resum the expansion .
the renormalisation group ( rg ) provides a consistent framework within which to carry this out .
as usual , the renormalised couplings @xmath184 are defined as the values at the normalisation point of the irreducible vertex functions @xmath185 , which are the truncated laplace - fourier transforms of @xmath186 .
( there is no field renormalisation in this theory . )
it is convenient to choose the normalisation point as @xmath187 this class of theories has the special property that the renormalised coupling constants may be computed to all orders@xcite .
the calculation is summarised in appendix a. the result is @xmath188 where @xmath189 it is convenient to define the dimensionless renormalised couplings as @xmath190 as will be seen , these are the natural expansion parameters for the renormalised perturbation expansion .
as discussed in the previous section , we are interested in computing the asymptotic behaviour at large @xmath3 of @xmath192 in the regularised bare theory , where @xmath193 . however , for the purposes of renormalising @xmath194 , it is more convenient to choose the time arguments of the fields @xmath195 to be independent , and to consider the laplace transform with respect to these times . thus we define @xmath196 and @xmath197 from this we may define the irreducible vertex function@xcite , by truncating the external propagators : @xmath198 the renormalised vertex function is @xmath199 where @xmath200 is fixed by the normalisation condition @xmath201 in writing ( [ eqn : ngamma ] ) , we have made it clear that that the ( un)renormalised vertex function is to be thought of as depending on the ( un)renormalised couplings . now ,
although we have used the condition ( [ eqn : z ] ) on @xmath202 to define @xmath200 , the _ same _ multiplicative renormalisation also renders @xmath203 finite , for @xmath204 , where @xmath205 is defined in ( [ eqn : overlineg ] ) . for this to be true , it is important that the fields @xmath206 are not evaluated at the same point .
this would lead to further uv divergences .
however , these occur on a set of measure zero in the integration in ( [ eqn : overlineg ] ) , and are harmless . _
_ define the rg functions @xmath208 and @xmath209 the fact that @xmath210 then implies the callan - symanzik equation @xmath211 if the couplings @xmath212 flow towards a nontrivial fixed point @xmath213 at which @xmath214 ( as we shall show happens for @xmath5 ) , then in estimating the leading asymptotic behaviour as @xmath215 it is sufficient to replace ( [ eqn : rgeq1 ] ) by @xmath216 where @xmath217 .
this has the solution @xmath218 , as @xmath215 at fixed @xmath3 .
however , simple dimensional analysis implies that @xmath219 is function of only the combination @xmath220 .
hence we find that @xmath221 with @xmath222 from ( [ ccr ] ) and ( [ dimless ] ) , we find after some algebra ( see app . a ) @xmath224 note that this is exact to all orders in @xmath225 , and that , for @xmath226 , there is an infrared stable fixed point at @xmath227 whose value is independent of the diffusion constants .
consider the expansion of the vertex function ( [ eqn : gamma ] ) as a power series in the coupling constants @xmath229 .
to first order , this is given by the sum of diagrams like that in fig .
[ fig1loop ] , explicitly @xmath230 where the integral @xmath231 is the same as occurs in the coupling constant renormalisation : see appendix a. the combinatorial factor @xmath232 counts the number of ways the walkers from family @xmath143 can interact just once with those of family @xmath144 .
the renormalisation constant @xmath200 is then the inverse of this evaluated at the normalisation point @xmath233 .
thus @xmath234 and so , by ( [ eqn : gamma ] ) @xmath235 next we set @xmath236 . by ( [ eqn : g * ] ) ,
@xmath237 through ( [ eqn : alpha ] ) , this gives the result ( [ eqn : result1 ] ) up to @xmath238 .
we remark that to this order the result is independent of the diffusion constants , as long as no pair @xmath239 vanishes .
this last is of course a pathological case , since then the two families are immobile and can not meet .
there are three types of diagrams contributing to @xmath240 at order @xmath241 .
they are illustrated in figs .
[ fig2loop]a - c .
contributions of types ( a ) and ( b ) involve the one - loop integrals @xmath242 defined above , while those of type ( c ) involve @xmath243 define @xmath244 then @xmath245 each term in this sum corresponds to a diagram of class @xmath246 to @xmath247 in fig .
[ fig2loop ] .
the combinatorial factors , polynomials in the @xmath14 , count the number of ways different walkers from a given family can contribute to each of these processes .
( it is simplest to check these factors for small values of @xmath2 and @xmath14 . ) from appendix a @xmath248 moreover , as shown in appendix b @xmath249 b_{d}^2 ( 2 \sigma)^{-\varepsilon } , \label{eqn : i2a}\end{aligned}\ ] ] where @xmath250 is given by ( [ eqn : bd ] ) and @xmath251 is given by ( [ eqn : r0 ] ) .
next we compute @xmath252 through @xmath253 , perform the differentiation @xmath254 at fixed bare couplings @xmath255 , then re - express the result as a series in the @xmath225 , to the same order .
the result has the form @xmath256 with @xmath257 an important check of this calculation is that the double poles in @xmath22 in the two - loop contributions are cancelled by the coupling constant renormalisation in the one - loop contributions .
this has the consequence that @xmath258 has a finite limit as @xmath259 , that is , @xmath260 vanishes .
this is shown in appendix c. now we set @xmath261 following ( [ eqn : g * ] ) .
then we have @xmath262 with @xmath263 through ( [ eqn : alpha ] ) , this gives the result ( [ eqn : result1 ] ) .
when @xmath51 the couplings @xmath264 are marginally irrelevant , that is they flow logarithmically slowly towards zero . in that case , it is not sufficient to set them equal to their fixed - point values , but instead the full callan - symanzik equation ( [ eqn : rgeq1 ] ) must be solved . using the fact that @xmath265 depends on @xmath266 and @xmath3 only through combination @xmath220 ,
this may be rewritten @xmath267 the solution by the method of characteristics is standard .
define running couplings @xmath268 by @xmath269 with initial conditions @xmath270 .
then @xmath271 in our case , @xmath272 so that , using ( [ gammao ] ) the exponent in ( [ eqn : cssoln ] ) is @xmath273 exponentiating this yields the result quoted in ( [ 2dlog ] )
. the dependence of the last factor in ( [ eqn : cssoln ] ) on @xmath274 also generates corrections which are down by @xmath275 .
all the non - universal behaviour resides in these , and higher order , corrections .
note the absence of corrections @xmath276 , which may be traced to the lack of higher - order terms in the beta - functions ( [ betafn ] ) .
we have presented a generalisation of the vicious walker problem in which walkers from different families annihilate on meeting , but walkers from the same family ignore each other .
we have studied the problem in a field - theoretic renormalisation group framework , suitable for understanding universal quantities such as critical exponents .
we have focussed on the probability that all walkers have survived up to time @xmath3 , and we have showed that , in dimension @xmath5 , this decays as a power , @xmath277 , where @xmath14 is the number of walkers in the @xmath13th family . while this result is true to all orders in @xmath54 , the actual values of the exponents can , by this method , be evaluated only as a power series in @xmath22 , which we have carried out to second order .
the coefficient of the @xmath278 term depends on the ratios of the diffusivities of each family , as well as the @xmath49 .
the lack of dependence of the @xmath238 term on this ratio may be traced mathematically to the fact that the same bubble diagram enters into the coupling constant renormalisation ( fig .
[ figcc ] ) and the renormalisation of the composite operator @xmath191 ( fig .
[ fig1loop ] ) . for the same reason , the exponent for @xmath279 walkers does not depend on the ratio of their diffusivities , because in this case the one - loop result is correct to all orders ( as long as we do not expand the coefficients in powers of @xmath22 . )
the same would be true , to first order for @xmath280 , and to all orders for @xmath281 , if the two - body interactions we consider were generalised to @xmath145-body interactions with @xmath282 , although in this case the upper critical dimension would be reduced to @xmath283 .
the value of @xmath53 was shown to be related to the anomalous dimension of a certain composite operator @xmath284 .
this structure means that our results can be straightforwardly extended to other physical observables .
for example , the _ reunion _ probability @xmath285 that all the walkers have survived up to time @xmath3 and are all located within a distance @xmath286 of each other ( where @xmath107 is for example the lattice spacing ) , is related to a correlation function @xmath287 where @xmath288 .
since the theory is symmetric under @xmath289 , @xmath290 has the same anomalous dimension as @xmath191 .
hence @xmath291 where the first term in the exponent comes from simple power counting , and the factor of 2 in the second reflects the important fact that the anomalous scaling of composite operators at different times ( and points ) is multiplicative . _
_ the fact that @xmath292 is symmetric under permutations of the families @xmath293 has the important consequence that the exponents @xmath53 are also symmetric .
the form of the @xmath22-expansion implies that this is true to all orders in @xmath22 .
it may be traced to the operator product expansion ( [ ope ] ) : the next - to - leading terms on the right - hand side , which do not vanish on integration over the spatial coordinates must contain at least two derivatives , for example @xmath294 power - counting shows that , at @xmath51 , the contribution of such terms is at least @xmath295 down on the leading term , and therefore , for sufficiently small @xmath22 , they yield only corrections to the leading behaviour which we have computed .
however , since each term in ( [ ope ] ) is renormalised separately , each gives rise to an independent scaling exponent .
it is a very interesting question whether the first term in the ope is dominant all the way down to @xmath23 .
such a result would imply that the asymptotic exponents ( but not necessarily the prefactors ) for the cases @xmath296 and @xmath297 , for example , are equal . yet , in one dimension , the two problems are certainly not isomorphic , because the ordering of the families along the real line makes a difference .
a very similar situation occurs in the problem of intersections of families of brownian paths in two dimensions : in that case the exponents have been computed exactly@xcite and are known to be symmetric , despite the fact that the ordering of the families around the annulus is a priori relevant . in this example , this symmetry is also suggested by the @xmath22-expansion below the upper critical dimension , in this case @xmath298 @xcite .
however , a recent result of bray and blythe@xcite suggests that the situation in @xmath23 is not so straightforward for our problem . in their eq .
( 4 ) they report a result for the survival exponent for a single lamb , with diffusion constant @xmath299 , with @xmath300 lions to the left and @xmath301 lions to the right .
the diffusion constant of the lions is @xmath302 , and their result is an expansion to first order in @xmath303 .
it is reported to depend on the asymmetry @xmath304 as well as the sum @xmath305 , while our analysis shows that it should depend only on the sum , to all orders in the @xmath22-expansion .
there are at least two possible resolutions : 1 .
the qualitative conclusions of the @xmath22-expansion break down in @xmath23 , either because the non - leading terms in the ope dominate , or through some more systemic failure ; 2 . for @xmath23
there is a qualitative difference between the case when walkers if different families strictly can not pass each other , and that , more appropriate to the field theory approach , when the annihilation rate is finite and therefore the order along the real line is not preserved . in any case , it may be shown , by generalising the arguments of krapivsky and redner@xcite , that for infinite annihilation rate in @xmath23 the exponents @xmath53 ( and indeed all the non - leading exponents ) , are simply related to the eigenvalues of the dirichlet problem in a certain compact region consisting of a spherical hyperpolygon on the sphere @xmath306 .
thus , if the postulated symmetry were to hold , it would imply that the laplacians in different regions which are related by permutations of the @xmath49 are isospectral .
we plan to address this question in a future publication . _
jc was supported in part by the engineering and physical sciences research council under grant gr / j78327 .
this work was done during a stay of mk in theoretical physics , oxford university .
mk would like to thank the department for hospitality .
_ _ _ 99 m. e. fisher , walks , walls , wetting , and melting , j. stat .
phys . * 34 * , 667 - 729 ( 1984 ) .
see also d. a. huse and m. e. fisher , commensurate melting , domain walls , and dislocations , phys .
b * 29 * , 239 - 270 ( 1984 ) .
s. mukherji and s. m. bhattacharjee , reunion and survival of interacting walkers , phys .
e * 48 * , 3427 - 3440 ( 1993 ) .
erratum , phys .
e * 52 * , 3301 ( 1995 ) .
s. mukherji and s. m. bhattacharjee , reunion of vicious walkers : results from @xmath307-expansion , j. phys .
a : math . gen . * 26 * , l1139-l1144 ( 1993 ) . c. krattenthaler , a. j. guttmann , and x. g. viennot , vicious walkers , friendly walkers and young tableaux : ii . with a wall , j. phys .
* 33 * , 8835 - 8866 ( 2000 ) .
m. katori and h. tanemura , scaling limit of vicious walks and two - matrix model , phys .
e * 66 * , 011105 ( 2002 ) .
de gennes , soluble model for fibrous structures with steric constraints , j. chem .
* 48 * , 2257 - 2259 ( 1968 ) .
l. krapivsky and s. redner , kinetics of a diffusion capture process : lamb besieged by a pride of lions , j. phys . a : math
* 29 * , 5347 - 5357 ( 1996 ) .
s. redner and p. l. krapivsky , capture of the lamb : diffusing predators seeking a diffusing prey , am .
* 67 * , 1277 - 1283 ( 1999 ) .
s. redner , _ a guide to first - passage processes _ , ch . 6 - 8 ( cambridge university press , 2001 ) .
h. niederhausen , the ballot problem with three candidates , europ . j. combinatorics * 4 * , 161 - 167 ( 1983 ) .
b. duplantier , intersections of random walks : direct renormalization approach , comm .
* 117 * , 279 - 329 ( 1988 ) .
b. duplantier and k .- h .
kwon , conformal invariance and intersections of random walks , phys .
lett . * 61 * , 2514 - 2517 ( 1988 ) .
g. f. lawler and w. werner , intersection exponents for planar brownian motion , ann .
* 27 * , 1601 - 1642 ( 1999 ) ; g. f. lawler , o. schramm and w. werner , sharp estimates for brownian non - intersection probabilities , math.pr/0101247 . m. doi , second quantization representation for classical many - particle system , j. phys .
a * 9 * , 1465 - 1477 , ( 1976 ) .
+ l. peliti , path integral approach to birth - death processes on a lattice , j. physique * 46 * , 1469 - 1483 ( 1985 ) .
+ j. cardy , field theory and nonequilibrium statistical mechanics , lecture notes ( 1999 ) , http://www-thphys.physics.ox.ac.uk/users/johncardy/notes.ps d. j. amit , _ field theory , the renormalization group , and critical phenomena _ ( world scientific , singapore , 1984 ) .
a. j. bray and r. a. blythe , exact asymptotics for one - dimensional diffusion with mobile traps , cond - mat/0205381 .
the diagrams contributing to @xmath308 are shown in fig .
[ figcc ] . in @xmath309 space
they give a geometric sum .
we have therefore the exact result @xmath310 where @xmath311 and @xmath312 .
the renormalised coupling is the value of this at @xmath313 and @xmath314 .
thus @xmath315 where @xmath316 the dimensionless coupling ( [ dimless ] ) is then given as @xmath317 thus @xmath318 differentiating this equation with respect to @xmath266 at fixed @xmath319 and using the definition of the beta - function ( [ eqn : beta ] ) @xmath320 that is , @xmath321
at the normalisation point , the two - loop diagram fig .
[ fig2loop]@xmath122 leads to the integral @xmath323 as usual , the denominators may be combined using a feynman parameter integration over @xmath324 : @xmath325 the wave - number integrals are now standard , and yield @xmath326 where @xmath327 this has a simple pole at @xmath51 , arising from the end - point at @xmath328 .
however , we also have to extract the finite part .
this may be done by writing @xmath329 where @xmath330\end{aligned}\ ] ] the second integral is finite at @xmath51 : @xmath331+{\cal o}(\varepsilon ) \nonumber\\ & = & -\ln r(d_1,d_2,d_3)+{\cal o}(\varepsilon)\end{aligned}\ ] ] where @xmath332 recalling the definition ( [ eqn : bd ] ) of @xmath333 , we therefore find , after some algebra , @xmath334\ ] ] which leads directly to ( [ eqn : i2a ] ) .
it should be remarked that a crucial simplification in this calculation arises because @xmath335
by ( [ eqn : b2a ] ) , @xmath337 | we consider a generalisation of the vicious walker problem in which @xmath0 random walkers in @xmath1 are grouped into @xmath2 families . using field - theoretic renormalisation group methods we calculate the asymptotic behaviour of the probability that no pairs of walkers from different families have met up to time @xmath3 . for @xmath4 ,
this is constant , but for @xmath5 it decays as a power @xmath6 , which we compute to @xmath7 in an expansion in @xmath8 .
the second order term depends on the ratios of the diffusivities of the different families . in two dimensions ,
we find a logarithmic decay @xmath9 , and compute @xmath10 exactly .
* families of vicious walkers * + 0.1 in john cardy + theoretical physics , 1 keble road , oxford ox1 3np , united kingdom + and all souls college , oxford + + department of physics , chuo university , kasuga , bunkyo - ku , tokyo 112 - 8551 , japan + |
inclusive unpolarized and polarized deeply inelastic diffractive scattering at high energies and momentum transfer is one of the important processes in lepton
nucleon scattering . as found by experiment , cf .
@xcite , there are interesting relations between the cross sections of these processes and those of inclusive deeply inelastic scattering : _ i ) _ the scaling violations of both processes are quite similar and _ ii ) _ the ratio of the differential cross sections in @xmath2 and @xmath3 are widely constant in the whole kinematic domain and are of @xmath4 . whereas the latter aspect can not be understood with perturbative methods the former calls for a rigorous analysis in perturbative qcd . in recent analyses @xcite
this aspect has been investigated both for the unpolarized and the polarized case on the basis of the light cone expansion . by this method
the semi - exclusive processes of diffractive scattering could be related to forward scattering processes at short distances , for which similar evolution equations as in the deep inelastic case apply .
moreover a callan gross and wandzura - wilczek relation between the twist2 contributions of the diffractive structure functions were derived . in this note
we give a summary of these papers .
the process of deep inelastic diffractive scattering is @xmath5 , with a significant rapidity gap between @xmath6 and the remaining hadrons .
the differential scattering cross section for single
photon exchange is given by @xmath7 with @xmath8 and @xmath9 the leptonic and hadronic tensors .
using current conservation , p and t invariance and the hermiticity relation for the hadronic tensor one finds a representation of the hadronic tensor in terms of four unpolarized and eight polarized structure functions @xcite .
we will henceforth consider the case of small values of @xmath10 . in this limit
the outgoing and incoming proton momenta are related by @xmath11 and the cross section depends on two unpolarized and two polarized structure functions only @xmath12 with @xmath13 and @xmath14 for @xmath15 .
( [ eqhadr ] ) is considered in the generalized bjorken limit : @xmath16 and @xmath17 = fixed .
the non - forward variable @xmath18 is related to another variable often used , @xmath19 , by @xmath20 . in the limit
@xmath21 the above structure functions depend on the three variables @xmath22 and @xmath3 . since for diffractive processes the outgoing proton
is well separated in rapidity from the diffractively produced hadrons ( rapidity gap ) , one may apply a. mueller s generalized optical theorem @xcite to calculate the scattering cross section .
this is done moving the outgoing proton into an incoming anti - proton and considering the absorptive part of deep inelastic forward scattering off the state @xmath23 summing over all final - state spins . note that under this interchange @xmath24 is kept space like . due to this operation
we may now evaluate the compton
operator @xmath25 \\ & = & -e^2 \frac{\tilde x^\lambda}{2 \pi^2 ( x^2-i\epsilon)^2 } rt \left [ \overline{\psi } \left(\frac{\tilde
x}{2}\right ) \gamma^\mu \gamma^\lambda \gamma^\nu \psi \left(-\frac{\tilde x}{2}\right ) - \overline{\psi } \left(-\frac{\tilde x}{2}\right ) \gamma^\mu \gamma^\lambda \gamma^\nu \psi \left(\frac{\tilde x}{2}\right ) \right ] s \nonumber\end{aligned}\ ] ] between the above states for forward scattering .
we represent this operator in terms of a vector and an axial - vector operator , which are in turn related to the associated scalar and pseudo - scalar operators , through which we introduce the respective operator expectation values , see @xcite defining non forward parton densities @xmath26 , @xmath27 with @xmath28 and @xmath29 . here
we neglect sub - leading components @xmath30 . after passing a series of steps , see @xcite , we may express the hadronic tensor in this approximation by one unpolarized and one polarized distribution function , @xmath31 and @xmath32 , respectively . for quarks and anti - quarks these distribution functions , which are the diffractive parton distributions , read @xmath33
the upper sign refers to quarks , the lower to anti - quarks , and @xmath34 in the unpolarized case , @xmath35 in the polarized case , where @xmath36 .
the diffractive structure functions @xmath37 and @xmath38 obey the representation @xmath39\nonumber\\ g_1^d(\beta,\eta , q^2 ) & = & \sum_{q=1}^{n_f } e_q^2 \left [ f_{q5}^d(\beta , x_{{{\mathbb{p}}}},q^2)+\overline{f}^d_{q5 } ( \beta , x_{{{\mathbb{p}}}},q^2)\right]~.\end{aligned}\ ] ] after some calculation one finds for the twist2 contributions to the hadronic tensor the relations @xmath40 the callan gross relation between the structure functions depending on @xmath41 is modified due to the emergence of @xmath2 , while the wandzura wilczek relation holds in the new variable @xmath42 $ ] .
the emergence of the integral term in one of the above relations is due to a basic connection between a vector valued non forward distribution function and the associated scalar one @xcite .
the corresponding term exceptionally cancels in the callan gross relation but is present in most relations of this type , see also @xcite .
the evolution equations of the diffractive parton densities can be formulated starting with the evolution equations for the scalar quark and gluon operators in the flavor non singlet and singlet case , see e.g. @xcite . @xmath43 with @xmath44 the factorization scale .
forming expectation values as in the foregoing section one notices that the evolution does not depend on the value of the light - cone mark @xmath45 , which can be set to 0 .
moreover the all - order rescaling relation @xmath46 where @xmath47 , is applied .
after some calculation one finds the following evolution equations @xmath48 these equations apply both to the unpolarized and polarized diffractive parton densities of twist2 to all orders in the coupling constant . in the calculation
the absorptive part of the distribution was taken , which identifies the original momentum fraction , cf .
@xcite , @xmath49 with @xmath50 and results in evolution equations with a support [ 0,1 ] , unlike those in @xmath51 . if compared to the case of deep - inelastic scattering the evolution is here not in @xmath2 but in the new variable @xmath41 .
otherwise the same evolution equations are obtained .
this holds also for higher twist operators , for which the argument is exactly the same as given in this section , see @xcite .
we derived the twist2 evolution equations for the unpolarized and polarized twist2 diffractive parton distributions considering these processes in the light
cone expansion at short distances .
we also derived relations between the diffractive structure functions in the unpolarized and polarized case .
the observed similarity of the scaling violations between deep - inelastic diffractive and deep
inelastic structure functions was shown in the present approach being due to the same structure of evolution equations which act in the former case on the variable @xmath41 and in the latter on @xmath2 . polarized deep
inelastic scattering was not yet observed nor studied in detail in a larger kinematic domain .
it would be interesting to know if the pattern of relations as observed for unpolarized scattering repeats and the predictions of the present paper can be verified .
the compass experiment and polarized experiments at future high - energy facilities @xcite could clarify this .
99 see e.g. h. abramowicz and j. dainton , j. phys .
* g 22 * ( 1996 ) 911 ; + j. breitweg et al . , zeus collaboration , eur .
j. * c6 * ( 1999 ) 43 ; + c. adloff et al . , h1 collaboration , z. phys . * c76 * ( 1997 ) 613 .
j. blmlein and d. robaschik , phys . lett .
* b517 * ( 2001 ) 222 .
j. blmlein and d. robaschik , phys.rev .
* d65 * ( 2002 ) 096002 .
mueller , phys . rev . * d2 * ( 1970 ) 2963 ; phys . rev . *
d4 * ( 1971 ) 150 ; + p.d.p .
collins , an introduction to regge theory and high energy physics , ( cambridge university press , cambridge , 1977 ) , pp . 331 .
j. blmlein , b. geyer , and d. robaschik , nucl . phys .
* b * nucl .
* b560 * ( 1999 ) 283 ; j. blmlein and d. robaschik , nucl . phys .
* b581 * ( 2000 ) 449 ; j. blmlein , j. eilers , b. geyer , and d. robaschik , phys . rev . * d65 * ( 2002 ) 054029 .
j. blmlein and n. kochelev , phys . lett .
* b381 * ( 1996 ) 296 ; nucl . phys .
* b498 * ( 1997 ) 285 .
j. blmlein and a. tkabladze , nucl . phys .
* b553 * ( 1999 ) 427 .
m. anselmino et al . , tesla - n study group : electron scattering with polarized targets at tesla desy 00160 , trp0020 , hep - ph/0011299 . | polarized inclusive deep - inelastic scattering is formulated in the light cone expansion .
the qcd evolution of the leading twist distribution functions is derived .
it is shown that the twist2 contribution to the structure functions @xmath0 is obtained via @xmath1 by a wandzura wilczek relation . |
we use shells made of vinylpolysiloxane ( a silicone - based elastomer ) with elastic moduli of @xmath0 , @xmath1 and @xmath2 mpa , as well as shells made of latex .
the shells range from @xmath3@xmath4 mm in diameter , @xmath3@xmath5 mm in length and approximately @xmath6@xmath0 mm in thickness .
we clamp one end of these shells onto a rigid nozzle and pass air through them at flow rates ranging from @xmath0@xmath7 liters per second .
when the flow rate of air in the shells exceeds a certain critical value , dependent on the dimensions and material properties of each shell , the shell becomes unstable and begins to vibrate .
the mode of vibration corresponds to one of the circumferential normal modes of vibration of cantilevered cylindrical shells .
which mode is observed depends on the dimensions and material properties of the shell .
we observed the first three modes . in the first mode , commonly known as the garden hose mode , " the shells oscillate side - to - side with the frequency of approximately @xmath8 hz . in the second mode , the surface of the shell bends inwards , obstructing the fluid flow and causing a large jump in the pressure drop across the nozzle . in this mode , the shell can vibrate with frequencies from @xmath9@xmath10 hz , depending on the volumetric flow rate of air .
we observed that the frequency of oscillation is directly proportional to air flow rate .
additionally , when the shell vibrates in the second mode with average frequencies ranging from approximately @xmath11@xmath12 hz , the vibration is unstable and the oscillation frequency varies widely between periods .
the second mode is the most robust and can be observed in the largest range of parameters in our shells . in the third mode of vibration , the circumference of the free end of the shell
is divided into three
flaps " oscillating inwards and outwards . in this mode ,
the shells vibrate with frequency of approximately @xmath13@xmath14 hz .
the images were captured with a phantom v5.2 high speed color camera .
the images of the second and third mode were captured using a stroboscopic technique and the final video is a concatenation of frames taken from different oscillation periods . each frame is slightly offset in phase , yielding a slow - motion effect .
we would like to express our gratitude to felice frankel for her help with cinematography , and to prof .
john bush who kindly lent us the high speed camera .
we would also like to thank prof .
pedro reis and dr .
arnaud lazarus for sharing with us their data on elastic properties of vinylpolysiloxane elastomers . | in this fluid dynamics video , we demonstrate the first three circumferential modes of fast , large amplitude vibrations of compliant cylindrical shells carrying a fluid . |
it has been known for a while that very intense electromagnetic fields are created in ultra - relativistic hadronic and nuclear collisions @xcite . however , no convincing experimental evidence of their impact on the scattering dynamics has been observed . in recent years , a renewed interest to this subject
was motivated by the relativistic heavy - ion collision experiments .
the electromagnetic fields are intense enough to modify the properties of the nuclear matter produced in these collisions . in order to evaluate the impact of these fields on the nuclear matter , it is crucial to know their space - time structure . in @xcite production of the electromagnetic fields
was studied using the hadron transport models , neglecting the nuclear medium electric and magnetic response and flow . in @xcite
it was pointed out that the quantum nature of the nucleus wave function gives rise to strong field fluctuation , so that even in central collisions the r.m.s .
does not vanish .
in @xcite it is argued that due to the finite electric conductivity of nuclear matter , the lifetime of the electromagnetic field is significantly longer than in vacuum .
anomalous transport can also affect the field producing oscillations @xcite and even forcing the field into the topologically non - trivial configurations @xcite . the electromagnetic field in the nuclear medium , unlike that in vacuum
, strongly depends on the initial conditions @xcite . the nuclear medium produced in relativistic heavy - ion collisions
is well described by the relativistic hydrodynamics .
relativistic magneto - hydrodynamic calculations were done in @xcite in the ideal limit ( infinite electrical conductivity ) . in a recent publication
@xcite we argued that one can treat the sources of the electromagnetic field , i.e. the valence quarks , neither as point particles ( classical limit ) nor as plane waves , which have infinite spatial extent .
this is because the interaction range , the quark wave function size and the dimensions of the produced nuclear matter have similar extent . as the first step towards understanding the quantum dynamics of the electromagnetic field sources , in @xcite we modeled valence quarks as spinless gaussian wave packets . solving the klein - gordon equation we computed the charge and current densities and the resulting electromagnetic fields in vacuum . in the present work
we extend our approach to compute the spin contribution to the electromagnetic field . as in @xcite
we start with the initial gaussian wave packet and evolve it in time according to the dirac equation . at this point
we completely neglect the medium effects as our goal is to study the effect of quantum diffusion of the quark wave function . this way our calculation is applicable to any hadronic and nuclear collisions .
before we proceed to the description of our calculation , it is worthwhile to set the notations , definitions and normalizations .
the wave function of a spin - half particle can be expanded in a complete set of the momentum and helicity eigenstates : [ a5 ] ( r , t ) = _
e^ik re^-i_k t _ k(0)u_k , where @xmath0 .
the four - component bispinor @xmath1 is the momentum and helicity eigenstate normalized as [ a7 ] u_k^u_k= _ . @xmath2 is the momentum wave function at @xmath3 , normalized as [ a9 ] |_k ( 0)|^2 d^3k= 1 . with these conventions [ a11 ] ^(r , t ) ( r , t ) d^3r = 1
. solutions of the dirac equation with given momentum @xmath4 and helicity @xmath5 normalized by ( [ a7 ] ) are [ a17 ] u_k + = ( c _ + + _
+ ) , u_k -= ( c _ - + _ - ) , where the two - component spinors @xmath6 are helicity eigenstates .
in the rest frame , although the particle momentum vanishes , the momentum of the fourier components in ( [ a5 ] ) is finite , which is the reason for the wave function diffusion .
although the particle spin projection on any axis is conserved , only spin projection on the momentum direction is conserved for states with given momentum .
this is why the helicity eigenstates are the correct choice of the spin basis . taking the direction of observation to be @xmath7-axis , i.e. @xmath8 and describing the momentum direction by the polar and azimuthal angles @xmath9 and
@xmath10 we write the helicity eigenstates [ b5 ] _ + = ( c + e^i ) , _ -= ( c + - e^i ) . using these in ( [ a17 ] ) yields [ b7 ] u_k + = ( c + e^i + + e^i ) , u_k -= ( c + -e^i + - + e^i ) , plugging ( [ b7 ] ) into ( [ a5 ] ) yields , after integration over the momentum directions ( keeping in mind that @xmath11 ) , the wave function in the rest frame [ b9 ] ( r , t)= _
0^dk k^2 e^-i_k t _ k ( 0 ) ( c f(kr ) + 0 + g(kr ) + 0 ) where f(z)&= _ -1 ^ 1(+)e^izxdx + & = \ { ( z)-(z ) s ( ) + ( z ) c ( ) } , [ b11 ] + g(z ) & = _
-1 ^ 1(-)e^izxdx + & = \ { -(z)+(z ) s ( ) + ( z ) c ( ) } .[b12 ] where @xmath12 and @xmath13 are the fresnel integrals related to the error function : [ b13 ] c(z)+is(z)= .
the corresponding charge and current densities are obtained using [ b14 ] = e^,j = e^. the resulting @xmath14 and @xmath15 are spherically symmetric . in particular @xmath15 is directed along the @xmath7-axis , which coincides with the direction of @xmath16 in our coordinate system .
[ b16 ] ( r , t)&= _
0^dk k^2 _ k ( 0 ) _ 0^dl l^2 _ l ( 0 ) + & \ { f(kr)f(lr)+ g(kr)g(lr ) } .
[ b18 ] j(r , t ) & = r _ 0^dk k^2 _ k ( 0 ) _
0^dl l^2 _ l ( 0 ) + & \ { f(kr)g(lr)- g(kr)f(lr ) } . given the charge and current densities ,
the electric field can be computed as [ b20 ] e(r , t)&= \ { + - } d^3r , where @xmath17 is the retarded time and @xmath18 .
the only non - vanishing component of the electromagnetic field in the rest frame is the radial component of the electric field @xmath19 : [ b22 ] & e ( r , t ) = e(r , t)r , & b(r , t ) = 0 .
we emphasize , that although the entire discussion of this section deals with a charged point particle at rest , its electromagnetic field @xmath20 is different from the coulomb field @xmath21 due to the quantum evolution of the particle wave function .
it is instructive to separate the convective and spin contributions ( marked below by the subscripts @xmath22 and @xmath23 respectively ) . using the gordon identity [ d5 ] |_2 ^_1=
-i_(|_2^_1 ) we can write the charge and current densities ( [ b14 ] ) as [ d7 ] = _ c+ _ s , j = j_c+j_s , where _ c&= ( |- ) , j_c= , [ d9 ] + _ s&= -(^),j_s= _ t(^)-(^_0 ) . [ d10 ] explicit expressions for the convective charge and current densities read [ d12 ]
_ c(r , t)&= _
0^dk k^2 _ k ( 0 ) _ 0^dl l^2 _ l ( 0 ) + & ( _ k+_l)\ { f(kr)f(lr)+ g(kr)g(lr ) } .
[ d13 ] j_c(r , t ) & = r _ 0^dk k^2 _ k ( 0 ) _
0^dl l^2 _ l ( 0 ) + & \{l f(kr)f(lr)-kf(kr)f(lr)+ } . here
the integrals f(z)&= -_-1 ^ 1(+)x(zx ) dx,[d15 ] + g(z)&= _ -1 ^ 1(-)x(zx ) dx[d16 ] can be expressed in terms of the fresnel integrals ( [ b13 ] ) .
the corresponding field is computed using ( [ b20 ] ) .
in the lab frame moving with constant velocity @xmath24 with respect to the rest frame , the electric charge moves with velocity @xmath25 .
in contrast to the previous sections , in this and the next sections , all quantities pertaining to the rest frame are denoted by the subscript @xmath26 , whereas those pertaining to the lab frame bear no such subscript .
the lorentz transformations relating the coordinates and the fields in the two frames are & r_0 = r _ , & z_0= ( z - vt),[c1 ] + & e_z0= e_z , & b_z0=b_z , [ c2 ] + & e_0= ( e_+ v zb _ ) , & b_0 = ( b_-v ze _ ) .
[ c3 ] here @xmath27 and @xmath28 .
thus , the electric field ( [ b22 ] ) in the lab frame transforms into & e_z= e_z0 , & b_z=0 , [ c5 ] + & e_= e_0 , & b_= v ze_0.[c6 ] using the cylindrical coordinates one can write [ c8 ] & e= e_b + e_zz , & b= b , where e_z(r , t)= e_0(,(t - vz ) ) , [ c12 ] + b(r , t)= e_0(,(t - vz ) ) , [ c13 ] and @xmath29 .
@xmath30 is the magnitude of the electric field in the rest frame .
impact parameter @xmath31 is a distance from the moving charge in the transverse direction .
to compute the electromagnetic field created by a relativistic wave packet , one needs to specify the initial wave function .
we adopted a gaussian distribution [ b3 ] _
k(0)= e^-a^2k^2/2 .
its width in the coordinate space is fixed at @xmath32 fm reflecting the strong interactions range .
we also assumed that the momentum wave function is independent of @xmath33 , i.e. it is the same for all spin states .
other parameters used in our numerical calculations were the valence quark mass @xmath34 gev , electric charge @xmath35 ( we omit the quark charges 2/3 and -1/3 ) and the boost factor @xmath36 . [ cols="^,^ " , ] the results of our calculations are exhibited in fig .
[ all]-fig .
[ bvt ] . shown in fig .
[ all ] is the dependence of the magnetic field on time @xmath37 ( left panel ) and distance @xmath31 ( right panel ) . on the left panel notice that at later time the magnetic field changes sign .
we have observed this effect before for a scalar particle @xcite , where it is , in fact , much more prominent .
we will add a few more comments about the sign flip later in this section . in fig .
[ bvb ] we plot each of the lines shown in fig . [ all ] ( left panel ) separately along with its convective and spin components .
also plotted is the corresponding classical ( boosted coulomb ) field for comparison . at large @xmath31 ,
the classical ( i.e. point ) and quantum ( i.e. wave packet ) sources induce the same field as expected .
while the convective current contributes to the monopole term falling off as @xmath38 at large @xmath31 , the leading spin current contribution starts with the dipole term which falls off as @xmath39 . at later times , due to the quantum diffusion
, the deviation from the classical field is observed in a wider range of distances . in fig .
[ bvt ] we plot each of the lines shown in fig . [ all ] ( right panel ) separately along with its convective and spin components as well as the classical field . in order to better demonstrate the sign flip dynamics we use the linear scale for the vertical axis .
it is seen that while the convective part of the magnetic field changes its sign from positive to negative , the spin part changes its sign from negative to positive at about the same time .
this makes the sign flip effect for the magnetic field generated by spin-1/2 particle less pronounced than for the field generated by a scalar particle ( computed in @xcite ) .
the origin of the spin flip can be traced back to ( [ b20 ] ) in which the last two terms are proportional to the rate of decrease of charge density and the corresponding increase of the current density due to the quantum diffusion of the wave packet in the outward radial direction .
we computed the magnetic field created by a single valence quark , which is represented by a wave packet satisfying the dirac equation and moving in free space with relativistic velocity .
we observed that the classical description of the valence quarks as point - particles is not accurate for calculations of the electromagnetic field in hadron and/or nuclear collisions as it breaks down at distances as large as @xmath40 fm at @xmath36 .
moreover it misses an important spin - flip effect that occurs due to the quantum diffusion of the wave packet .
this paper is a step towards our ultimate goal of determining the structure and dynamics of electromagnetic fields created in relativistic hadron and nuclear collisions , and in particular , in heavy - ion collisions . in the forthcoming publication
, we intend to study the effect of the conducting medium on the electromagnetic field created by quantum sources .
w. greiner , b. muller and j. rafelski , `` quantum electrodynamics of strong fields , '' berlin , germany : springer ( 1985 ) 594 p. ( texts and monographs in physics ) j. rafelski and b. muller , `` magnetic splitting of quasimolecular electronic states in strong fields , '' phys . rev .
lett . * 36 * , 517 ( 1976 ) .
j. ambjorn and p. olesen , `` on electroweak magnetism , '' nucl .
b * 315 * , 606 ( 1989 ) .
j. ambjorn and p. olesen , `` w condensate formation in high - energy collisions , '' phys .
b * 257 * , 201 ( 1991 ) .
p. olesen , `` w condensation at the lhc ?
, '' arxiv:1207.7045 [ hep - ph ] .
d. e. kharzeev , l. d. mclerran and h. j. warringa , `` the effects of topological charge change in heavy ion collisions : event by event p and cp violation , '' nucl .
a * 803 * , 227 ( 2008 ) .
v. skokov , a. y. illarionov and v. toneev , `` estimate of the magnetic field strength in heavy - ion collisions , '' int .
j. mod .
phys . a * 24 * , 5925 ( 2009 ) v. voronyuk , v. d. toneev , w. cassing , e. l. bratkovskaya , v. p. konchakovski and s. a. voloshin , `` ( electro-)magnetic field evolution in relativistic heavy - ion collisions , '' phys . rev .
c * 83 * , 054911 ( 2011 ) j. bloczynski , x. g. huang , x. zhang and j. liao , `` azimuthally fluctuating magnetic field and its impacts on observables in heavy - ion collisions , '' phys .
b * 718 * , 1529 ( 2013 ) l. ou and b. a. li , `` magnetic effects in heavy - ion collisions at intermediate energies , '' phys . rev .
c * 84 * , 064605 ( 2011 ) w. t. deng and x. g. huang , `` event - by - event generation of electromagnetic fields in heavy - ion collisions , '' phys .
c * 85 * , 044907 ( 2012 ) a. bzdak and v. skokov , `` event - by - event fluctuations of magnetic and electric fields in heavy ion collisions , '' phys .
b * 710 * , 171 ( 2012 ) k. tuchin , `` synchrotron radiation by fast fermions in heavy - ion collisions , '' phys .
c * 82 * , 034904 ( 2010 ) [ erratum - ibid .
c * 83 * , 039903 ( 2011 ) ] .
b. g. zakharov , `` electromagnetic response of quark - gluon plasma in heavy - ion collisions , '' phys
. lett .
b * 737 * , 262 ( 2014 ) k. tuchin , `` time and space dependence of the electromagnetic field in relativistic heavy - ion collisions , '' phys
. rev .
c * 88 * , no . 2 , 024911 ( 2013 ) k. tuchin , `` particle production in strong electromagnetic fields in relativistic heavy - ion collisions , '' adv .
high energy phys . * 2013 * , 490495 ( 2013 ) k. tuchin , `` electromagnetic field and the chiral magnetic effect in the quark - gluon plasma , '' phys .
c * 91 * , no .
6 , 064902 ( 2015 ) c. manuel and j. m. torres - rincon , `` dynamical evolution of the chiral magnetic effect : applications to the quark - gluon plasma , '' phys .
d * 92 * , no . 7 , 074018 ( 2015 ) h. li , x. l. sheng and q. wang , `` electromagnetic fields with electric and chiral magnetic conductivities in heavy ion collisions , '' phys . rev .
c * 94 * , no . 4 , 044903 ( 2016 ) k. tuchin , `` excitation of chandrasekhar - kendall photons in quark gluon plasma by propagating ultrarelativistic quarks , '' phys .
c * 93 * , no . 5 , 054903 ( 2016 ) k. tuchin , `` spontaneous topological transitions of electromagnetic fields in spatially inhomogeneous cp - odd domains , '' phys .
c * 94 * , no .
6 , 064909 ( 2016 ) y. hirono , d. e. kharzeev and y. yin , `` quantized chiral magnetic current from reconnections of magnetic flux , '' phys .
* 117 * , no .
17 , 172301 ( 2016 ) y. hirono , d. kharzeev and y. yin , `` self - similar inverse cascade of magnetic helicity driven by the chiral anomaly , '' phys .
d * 92 * , no .
12 , 125031 ( 2015 ) x. l. xia , h. qin and q. wang , `` approach to chandrasekhar - kendall - woltjer state in a chiral plasma , '' phys .
d * 94 * , no .
5 , 054042 ( 2016 ) z. qiu , g. cao and x. g. huang , `` on electrodynamics of chiral matter , '' phys .
d * 95 * , no . 3 , 036002 ( 2017 ) k. tuchin , `` initial value problem for magnetic fields in heavy ion collisions , '' phys .
c * 93 * , no . 1 , 014905 ( 2016 ) g. inghirami , l. del zanna , a. beraudo , m. h. moghaddam , f. becattini and m. bleicher , `` numerical magneto - hydrodynamics for relativistic nuclear collisions , '' eur . phys .
j. c * 76 * , no .
12 , 659 ( 2016 ) v. roy , s. pu , l. rezzolla and d. rischke , phys .
b * 750 * , 45 ( 2015 ) s. pu , v. roy , l. rezzolla and d. h. rischke , `` bjorken flow in one - dimensional relativistic magnetohydrodynamics with magnetization , '' phys .
d * 93 * , no . 7 , 074022 ( 2016 ) | we compute electromagnetic fields created by a relativistic charged spin - half particle in empty space at distances comparable to the particle compton wavelength .
the particle is described as a wave packet evolving according to the dirac equation .
it produces the electromagnetic field that is essentially different from the coulomb field due to the quantum diffusion effect . |
diffusive particle acceleration at non - relativistic shock fronts is an extensively studied phenomenon .
detailed discussions of the current status of the investigations can be found in some excellent reviews ( drury 1983 ; blandford & eichler 1987 ; berezhko & krimsky 1988 ; jones & ellison 1991 ; malkov & drury 2001 ) . while much is by
now well understood , some issues are still subjects of much debate , for the theoretical and phenomenological implications that they may have .
one of the most important of these is the reaction of the accelerated particles on the shock : the violation of the _ test particle approximation _ occurs when the acceleration process becomes sufficiently efficient that the pressure of the accelerated particles is comparable with the incoming gas kinetic pressure .
both the spectrum of the particles and the structure of the shock are changed by this phenomenon , which is therefore intrinsically nonlinear . at present
there are three viable approaches to determine the reaction of the particles upon the shock : one is based on the ever - improving numerical simulations ( jones & ellison 1991 ; bell 1987 ; ellison , mbius & paschmann 1990 ; ellison , baring & jones 1995 , 1996 ; kang & jones 1997 ; kang , jones & gieseler 2002 ; kang & jones 2005 ) that allow one to achieve a self - consistent treatment of several effects .
the second approach is based on the so - called two - fluid model , and treats cosmic rays as a relativistic second fluid .
this class of models was proposed and discussed in ( drury & vlk 1980 , 1981 ; drury , axford & summers 1982 ; axford , leer & mckenzie 1982 ; duffy , drury & vlk 1994 ) .
these models allow one to obtain the thermodynamics of the modified shocks , but do not provide information about the spectrum of accelerated particles .
the third approach is semi - analytical and may be very helpful to understand the physics of the nonlinear effects in a way that sometimes is difficult to achieve through simulations , due to their intrinsic complexity and limitations in including very different spatial scales .
blandford ( 1980 ) proposed a perturbative approach in which the pressure of accelerated particles was treated as a small perturbation . by construction
this method provides the description of the reaction only for weakly modified shocks .
alternative approaches were proposed by eichler ( 1984 ) , ellison & eichler ( 1984 ) , eichler ( 1985 ) and ellison & eichler ( 1985 ) , based on the assumption that the diffusion of the particles is sufficiently energy dependent that different parts of the fluid are affected by particles with different energies .
the way the calculations are carried out implies a sort of separate solution of the transport equation for subrelativistic and relativistic particles , so that the two spectra must be somehow connected at @xmath0 _ a posteriori_. in ( berezhko , yelshin & ksenofontov 1994 ; berezhko , ksenofontov & yelshin 1995 ; berezhko 1996 ) the effects of the non - linear reaction of accelerated particles on the maximum achievable energy were investigated , together with the effects of geometry .
the maximum energy of the particles accelerated in supernova remnants in the presence of large acceleration efficiencies was also studied by ptuskin & zirakashvili ( 2003a , b ) .
the need for a _
practical _ solution of the acceleration problem in the non - linear regime was recognized by berezhko & ellison ( 1999 ) , where a simple analytical broken - power - law approximation of the non - linear spectra was presented .
recently , some promising analytical solutions of the problem of non - linear shock acceleration have appeared in the literature ( malkov 1997 ; malkov , diamond & vlk 2000 ; blasi 2002 , 2004 ) .
blasi ( 2004 ) considered for the first time the important effect of seed pre - existing particles in the acceleration region ( the linear theory of this phenomenon was first studied by bell ( 1978 ) ) .
in a recent work by kang & jones ( 2005 ) the seed particles were included in numerical simulations of the acceleration process .
numerical simulations have been instrumental to identify the dramatic effects of the particles reaction : they showed that even when the fraction of particles injected from the thermal gas is relatively small , the energy channelled into these few particles can be an appreciable part of the kinetic energy of the unshocked fluid , making the test particle approach unsuitable .
the most visible effects of the reaction of the accelerated particles on the shock appear in the spectrum of the accelerated particles , which shows a peculiar hardening at the highest energies .
the analytical approaches reproduce well the basic features arising from nonlinear effects in shock acceleration .
there is an important point which is still lacking in the calculations of the non - linear particle acceleration at shock waves , namely the possible amplification of the background magnetic field , found in the numerical simulations by lucek & bell ( 2000 , 2000a ) and bell & lucek ( 2001 ) and recently described by bell ( 2004 ) .
this effect is still ignored in all calculations of the reaction of cosmic rays on the shock structure .
we will not include this effect in the present paper .
nonlinear effects in shock acceleration of thermal particles result in the appearance of multiple solutions in certain regions of the parameter space .
this phenomenon is very general and was found in both the two - fluid models ( drury & vlk 1980 , 1981 ) and in the kinetic models ( malkov 1997 ; malkov et al .
2001 ; blasi 2004 ) .
monte carlo approaches do not show multiple solutions .
this behaviour resembles that of critical systems , with a bifurcation occurring when some threshold is reached in a given order parameter . in the case of shock acceleration , it is not easy to find a way of discriminating among the multiple solutions when they appear . in ( mond & drury 1998 )
, a two fluid approach has been used to demonstrate that when three solutions appear , the one with intermediate efficiency for particle acceleration is unstable to corrugations in the shock structure and emission of acoustic waves .
plausibility arguments may be put forward to justify that the system made of the shock plus the accelerated particles may sit at the critical point ( see for instance the paper by malkov , diamond & vlk ( 2000 ) ) , but we are not aware of any real proof that this is what happens . the physical parameters that play a role in this approach to criticality are the maximum momentum achievable by the particles in the acceleration process , the mach number of the shock , and the injection efficiency , namely the fraction of thermal particles crossing the shock that are accelerated to nonthermal energies .
the last of them , the injection efficiency , hides a crucial physics problem by itself , and plays an important role in establishing the level of shock modification .
this efficiency parameter in reality is determined by the microphysics of the shock and should not be a free parameter of the problem .
unfortunately , our poor knowledge of such microphysics , in particular for collisionless shocks , does not allow us to establish a clear and universal connection between the injection efficiency and the macroscopic shock properties .
put aside the possibility to have a fully self - consistent picture of this phenomenon , one can try to achieve a phenomenological description of it .
kang , jones & gieseler ( 2002 ) introduced a sort of weight function to determine a return probability of particles in the downstream fluid to the upstream fluid , as a function of particle momentum .
only sufficiently suprathermal particles can jump back to the upstream region and therefore take part in the acceleration process .
here we adopt an injection recipe which is similar to the _ thermal leakage _ model of kang et al .
( 2002 ) ( see also previous papers by malkov ( 1998 ) and by gieseler et al .
( 2000 ) ) and implement it in the semi - analytical approach of blasi ( 2002 , 2004 ) .
we investigate then the phenomenon of multiple solutions and show that the injection recipe dramatically reduces the appearance of these situations .
we also study in some detail the efficiency for particle acceleration as a function of the mach number of the shock and the maximum momentum of the accelerated particles .
the paper is structured as follows : in section [ sec : nonlin ] we briefly describe the method proposed by blasi ( 2002 ) for the calculation of the spectum and pressure of particles accelerated at a modified shock .
we describe the appearance of multiple solutions in section [ sec : multiplesolutions ] , and the comparison with the method of malkov ( 1997 ) in section [ sec : comparisonmalkov ] . in section [ sec : injectionrecipe ] we introduce a recipe for the injection of particles from the thermal pool .
this recipe is then used in section [ sec : singlesolutions ] to show how the regions of parameter space where multiple solutions appear are drastically reduced by the self - regulated injection . in section [ sec : escape ] we discuss the efficiency of particle acceleration at modified shocks , and stress the role of escape of particles from upstream infinity .
the consequences of the cosmic ray modification on the shock heating are investigated in section [ sec : heating ] .
we conclude in section [ sec : conclusions ] .
in this section we describe the method proposed by blasi ( 2002 , 2004 ) for the calculation of the spectrum and pressure of the particles accelerated at a shock surface , when the reaction of the particles is taken into account .
no seed particles are included here .
the equation that describes the diffusive transport of particles in one dimension is @xmath1 - u \frac{\partial f ( x , p)}{\partial x } + \ ] ] @xmath2 where we assumed stationarity ( @xmath3 ) .
the @xmath4 axis is oriented from upstream to downstream , as in fig .
the pressure of the accelerated particles slows down the fluid upstream before it crosses the shock surface , therefore the gas velocity at upstream infinity , @xmath5 , is different from @xmath6 , the fluid speed just upstream of the shock .
the injection term is taken in the form @xmath7 . as a first step ,
we integrate eq .
[ eq : trans ] around @xmath8 , from @xmath9 to @xmath10 , denoted in fig . 1 as points `` 1 '' and `` 2 '' respectively , so that the following equation can be written : @xmath11_2 - \left [ d \frac{\partial f}{\partial
x}\right]_1 + \frac{1}{3 } p \frac{d f_0}{d p } ( u_2 - u_1 ) + q_0(p)= 0,\ ] ] where @xmath6 ( @xmath12 ) is the fluid speed immediately upstream ( downstream ) of the shock and @xmath13 is the particle distribution function at the shock location . by requiring that the distribution function downstream is independent of the spatial coordinate ( homogeneity ) , we obtain @xmath14_2=0 $ ] , so that the boundary condition at the shock can be rewritten as @xmath11_1 = \frac{1}{3 } p \frac{d f_0}{d p } ( u_2 - u_1 ) + q_0(p ) .
\label{eq : boundaryshock}\ ] ] we can now perform the integration of eq .
[ eq : trans ] from @xmath15 to @xmath9 ( point `` 1 '' ) , in order to take into account the boundary condition at upstream infinity . using eq .
[ eq : boundaryshock ] we obtain @xmath16 @xmath17 we introduce the quantity @xmath18 defined as @xmath19 whose physical meaning is instrumental to understand the nonlinear reaction of particles .
the function @xmath18 is the average fluid velocity experienced by particles with momentum @xmath20 while diffusing upstream away from the shock surface . in other words ,
the effect of the average is that , instead of a constant speed @xmath6 upstream , a particle with momentum @xmath20 experiences a spatially variable speed , due to the pressure of the accelerated particles . since the diffusion coefficient is in general @xmath20-dependent , particles with different energies _ feel _ a different compression coefficient , higher at higher energies
if , as expected , the diffusion coefficient is an increasing function of momentum ( see ( blasi 2004 ) for further details on the meaning of the quantity @xmath18 ) . with the introduction of @xmath18 , eq .
[ eq : step ] becomes @xmath21 + q_0(p ) = 0 , \label{eq : step1}\ ] ] where we used the fact that @xmath22.\ ] ] the solution of eq .
[ eq : step1 ] for a monochromatic injection at momentum @xmath23 is @xmath24 @xmath25\right\}=\ ] ] @xmath26 @xmath27\right\}. \label{eq : inje}\ ] ] here we used @xmath28 , with @xmath29 the gas density immediately upstream ( @xmath9 ) and @xmath30 the fraction of the particles crossing the shock which take part in the acceleration process .
here we introduced the two quantities @xmath31 and @xmath32 , which are respectively the compression factor at the gas subshock and the total compression factor between upstream infinity and downstream . for a modified shock
, @xmath33 can attain values much larger than @xmath34 and more in general , much larger than @xmath35 , which is the maximum value achievable for an ordinary strong non - relativistic shock .
the increase of the total compression factor compared with the prediction for an ordinary shock is responsible for the peculiar flattening of the spectra of accelerated particles that represents a feature of nonlinear effects in shock acceleration . in terms of @xmath34 and
@xmath33 the density immediately upstream is @xmath36 . in eq .
[ eq : inje ] we can introduce a dimensionless quantity @xmath37 so that @xmath38 @xmath39 the nonlinearity of the problem reflects in the fact that @xmath40 is in turn a function of @xmath13 as it is clear from the definition of @xmath18 . in order to solve the problem we need to write the equations for the thermodynamics of the system including the gas , the cosmic rays accelerated from the thermal pool and the shock itself .
the velocity , density and thermodynamic properties of the fluid can be determined by the mass and momentum conservation equations , with the inclusion of the pressure of the accelerated particles .
we write these equations between a point far upstream ( @xmath15 ) , where the fluid velocity is @xmath5 and the density is @xmath41 , and the point where the fluid velocity is @xmath18 ( density @xmath42 ) .
the index @xmath20 denotes quantities measured at the point where the fluid velocity is @xmath18 , namely at the point @xmath43 that can be reached only by particles with momentum @xmath44 ( this is clearly an approximation , but as shown in section [ sec : comparisonmalkov ] it provides a good agreement with other calculations where this approximation is not used ) .
the mass conservation implies : @xmath45 conservation of momentum reads : @xmath46 where @xmath47 is the gas pressure at the point @xmath48 and @xmath49 is the pressure of accelerated particles at the same point ( we use the symbol @xmath50 to mean _ cosmic rays _ , in the sense of _ accelerated particles _ ) .
the mass and momentum escaping fluxes in the form of accelerated particles have reasonably been neglected .
note that at this point the equation for energy conservation has not been used .
our basic assumption , similar to that used in ( eichler 1984 ) , is that the diffusion is @xmath20-dependent and more specifically that the diffusion coefficient @xmath51 is an increasing function of @xmath20 .
therefore the typical distance that a particle with momentum @xmath20 travels away from the shock is approximately @xmath52 , larger for high energy particles than for lower energy particles increases with @xmath20 faster than @xmath18 does , therefore @xmath53 is a monotonically increasing function of @xmath20 . ] . as a consequence , at each given point @xmath43 only particles with momentum larger than @xmath20 are able to affect appreciably the fluid . strictly speaking the validity of the assumption depends on
how strongly the diffusion coefficient depends on the momentum @xmath20 ( see section [ sec : comparisonmalkov ] ) .
since only particles with momentum @xmath44 can reach the point @xmath48 , we can write @xmath54 where @xmath55 is the velocity of particles with momentum @xmath20 , @xmath56 is the maximum momentum achievable in the specific situation under investigation . from eq .
[ eq : pressure ] we can see that there is a maximum distance , corresponding to the propagation of particles with momentum @xmath56 such that at larger distances the fluid is unaffected by the accelerated particles and @xmath57 .
the equation for momentum conservation is : @xmath58 + \frac{1}{\rho_0 u_0 ^ 2 } \frac{d p_{cr}}{dp } = 0.\ ] ] using the definition of @xmath59 and multiplying by @xmath20 , this equation becomes @xmath60 = \frac{4\pi}{3 \rho_0 u_0 ^ 2 } p^4
v(p ) f_0(p ) , \label{eq : eqtosolve}\ ] ] where @xmath13 is known once @xmath40 is known .
[ eq : eqtosolve ] is therefore an integral - differential nonlinear equation for @xmath40 .
the solution of this equation also provides the spectrum of the accelerated particles .
the last missing piece is the connection between @xmath34 and @xmath33 , the two compression factors appearing in eq .
[ eq : inje1 ] .
the compression factor at the gas shock around @xmath8 can be written in terms of the mach number @xmath61 of the gas immediately upstream through the well known expression @xmath62 on the other hand , if the upstream gas evolution is adiabatic , then the mach number at @xmath9 can be written in terms of the mach number of the fluid at upstream infinity @xmath63 as @xmath64 so that from the expression for @xmath34 we obtain @xmath65^{\frac{1}{\gamma_g+1}}. \label{eq : rsub_rtot}\ ] ] now that an expression between @xmath34 and @xmath33 has been found , eq .
[ eq : eqtosolve ] basically is an equation for @xmath34 , with the boundary condition that @xmath66 . finding the value of @xmath34 ( and the corresponding value for @xmath33 ) such that @xmath66 also provides the whole function @xmath40 and , through eq .
[ eq : inje1 ] , the distribution function @xmath67 .
if the reaction of the accelerated particles is small , the _ test particle _ solution is recovered .
in the problem described in the previous section there are several independent parameters . while the mach number of the shock and the maximum momentum of the particles are fixed by the physical conditions in the environment , the injection momentum and the acceleration efficiency are free parameters .
the procedure to be followed to determine the solution was defined in ( blasi 2002 ) : the basic problem is to find the value of @xmath34 ( and therefore of @xmath33 ) for which @xmath66 . in fig .
2 we plot @xmath68 as a function of @xmath33 , for @xmath69 , @xmath70 and @xmath71 in the left panel and @xmath72 in the right panel ( @xmath73 here is the mass of protons ) .
the parameter @xmath30 was taken @xmath74 in the left panel and @xmath75 in the right panel .
the different curves refer to different choices of the mach number at upstream infinity .
the physical solutions are those corresponding to the intersection points with the horizontal line @xmath66 , so that multiple solutions occur for those values of the parameters for which there is more than one intersection with @xmath66 .
these solutions are all physically acceptable , as far as the conservation of mass , momentum and energy are concerned .
[ fig : sol ] it can be seen from both panels in fig .
2 that for low values of the mach number , only one solution is found . this solution may be significantly far from the quasi - linear solution .
indeed , for @xmath76 the solution corresponds to @xmath77 , instead of the usual @xmath78 solution expected in the linear regime .
lower values of the mach number are required to fully recover the linear solution .
when the mach number is increased , there is a threshold value for which three solutions appear , one of which is the quasi - linear solution .
for very large values of the mach number the solution becomes one again , and it coincides with the quasi - linear solution . [
fig : multi2 ] in fig .
3 we show the appearance of the multiple solutions for the case @xmath69 , @xmath70 and @xmath72 with mach number @xmath79 ( @xmath71 and @xmath79 ) in the left ( right ) panel .
the curves here are obtained by changing the value of @xmath30 .
the same comments we made for fig . 2 apply here as well : low values of @xmath30 correspond to weakly modified shocks , while for increasingly larger efficiencies multiple solutions appear .
the solution becomes one again in the limit of large efficiencies , and it always corresponds to strongly modified shocks .
the problem of multiple solutions is not peculiar of the kinetic approaches to the non - linear theories of particle acceleration at shock waves .
the same phenomenon was in fact found initially in two - fluid models ( drury & vlk 1980 , 1981 ) , where however no information on the spectrum of the accelerated particles and on the injection efficiency was available .
multiple solutions were also found by malkov ( 1997 ) and malkov et al .
( 2000 ) , in the context of a semi - analytical kinetic approach .
aside from the technical differencies between that method and the one proposed by blasi ( 2002 , 2004 ) , the main difference is in the fact that the former requires the knowledge of the exact expression for the diffusion coefficient as a function of the momentum of the particles , while the latter only requires that such diffusion coefficient is an increasing function of the particle momentum .
while the first approach may provide us with an _ exact _ solution to the problem , the second is in fact more practical , in the sense of providing an approximate solution even in those cases , the majority , in which no detailed information on the diffusion properties of the fluid is available .
the solution provided in ( blasi 2002 , 2004 ) is particularly accurate when the diffusion coefficient is bohm - like , @xmath80 , expected in the case of saturated self - generation of waves in the vicinity of the shock surface ( lagage & cesarsky 1983 ) .
we will now discuss the quantitative comparison between the results of malkov ( 1997 ) and those of blasi ( 2002 , 2004 ) , by considering a single situation in which multiple solutions are predicted ( in both approaches ) , and determining the spectra and compression factors in both methods .
we start with briefly summarizing the approach of malkov ( 1997 ) .
the following flow potential is introduced there : @xmath81 which is used as a new independent spatial variable to replace @xmath4 . using the flow potential
, it is possible to define an integral transformation of the flow profile as follows : @xmath82 \frac{du}{dx } dx,\ ] ] where @xmath83 is the spectral index of the particle distribution function and @xmath51 is the diffusion coefficient , which is assumed to be independent of the position .
an integral equation for @xmath84 can be derived by applying eq .
[ inttra ] to the @xmath4derivative of the euler equation ( malkov 1997 ; malkov et al .
2000 ) : @xmath85 @xmath86^{-1 } \times\ ] ] @xmath87,\ ] ] where @xmath88 is an injection parameter defined as @xmath89 and related to the compression factor by the following equation ( malkov 1997 ) : @xmath90 @xmath91\bigg\}^{-1}. \label{mal2}\ ] ] here @xmath92 is the compression factor in the shock precursor .
[ eq : rsub_rtot ] , [ mal1 ] and [ mal2 ] form a closed system that can be solved numerically . before showing the results it is worth noticing that the injection parameters @xmath30 and @xmath88 adopted in the two approaches are defined in two non equivalent ways .
however , the relation between @xmath30 and @xmath88 can easily be found by using eqs .
[ eq : inje1 ] and [ efficiency ] and can be written as : @xmath93 one may notice that this relation contains the compression factors @xmath33 and @xmath34 , which are what we are searching for .
this fact implies that three solutions characterized by the same value of @xmath30 may correspond to three distinct values of @xmath88 . in order to compare the results of the two different approaches we consider a shock having mach number @xmath94 and temperature at upstream infinity @xmath95 .
we set the value of the injection and maximum momenta equal to @xmath96 and @xmath97 respectively and we adopt an efficiency @xmath98 . using the approach proposed by blasi ( 2002 ) , we find three solutions , characterized by the values of the compression factor @xmath99 .
the last solution is the quasi - linear one , in which the precursor is very weak .
we adopt now these three values for the precursor compression factor to solve the system of equations given by eq .
[ eq : rsub_rtot ] , [ mal1 ] , and [ mal2 ] ) for different choices of the diffusion coefficient . in fig .
[ fig : malk_vel ] we plot the velocity profiles for the three solutions , as derived with the method of blasi ( 2002 , 2004 ) and detailed above ( solid line ) . in the figure @xmath37 with @xmath18 defined through eq .
[ eq : up ] for the method of blasi ( 2002 , 2004 ) .
it is easy to show that @xmath40 is related to the @xmath84 through the relation @xmath100 .
the dotted and dashed lines are the results obtained with the calculation of malkov ( 1997 ) with a bohm and kolmogorov diffusion respectively . for bohm diffusion the two approaches give very similar results .
for kolmogorov diffusion the differencies are larger , as expected . in fig .
[ fig : malk_spec ] we plot the spectra of the accelerated particles , as obtained in this paper ( solid line ) and for a bohm ( dotted line ) and kolmogorov ( dashed line ) diffusion coefficient , as derived by carrying out the calculation of malkov ( 1997 ) .
we recall that from the theoretical point of view the bohm diffusion is in fact what should be expected in the proximity of a shock if the turbulence necessary for the acceleration is strong and generated by the same cosmic rays that are being accelerated ( lagage & cesarsky 1983 ) . in this perspective
, we look at the results illustrated in this section as very encouraging in using the approach presented in blasi ( 2002 , 2004 ) , since it is simple and at the same time accurate in reproducing the major physical aspects of particle acceleration at cosmic ray modified shocks .
the presence of multiple solutions is typical of many non - linear problems and should not be surprising from the mathematical point of view . in terms of physical understanding
however , multiple solutions may be disturbing .
the typical situation that takes place in nature when multiple solutions appear in the description of other non linear systems is that ( at least ) one of the solutions is unstable and the system _ falls _ in a stable solution when perturbed .
the stable solutions are the only ones that are physically meaningful .
some attempts to investigate the stability of cosmic ray modified shock waves have been made by mond & drury ( 1998 ) and toptygin ( 1999 ) , but all of them refer to the two - fluid models .
a step forward is being carried out by blasi & vietri ( in preparation ) in the context of kinetic models .
in addition to the stability , another issue that enters the physical description of our problem is the identification of possible processes that determine some type of backreaction on the system .
it may be expected that when some types of processes of self - regulation are included , the phenomenon of multiple solutions is reduced . in this section
we investigate the type of reaction that takes place when a self - consistent , though simple , recipe for the injection of particles from the thermal pool is adopted .
this recipe is similar to that proposed by kang , jones & gieseler ( 2002 ) in terms of the underlying physical interpretation of the injection , but probably simpler in its implementation . for non - relativistic shocks
, the distribution of particles downstream is quasi - isotropic , so that the flux of particles crossing the shock surface from downstream to upstream can be written as @xmath101 , \label{eq : flux1}\ ] ] where @xmath55 is the velocity of particles with momentum @xmath20 and @xmath102 is the shock speed in the frame comoving with the downstream fluid .
the term @xmath103 is the component along the direction perpendicular to the shock surface of the velocity of particles with momentum @xmath20 moving in the direction @xmath104 .
it follows that the flux of particles moving tangent to the shock surface ( namely with @xmath105 ) is zero .
we recall that , having in mind collisionless shocks , the typical thickness of the shock , @xmath106 , is the collision length associated with the magnetic interactions that give rise to the formation of the discontinuity .
useless to say that these interactions are all but well known , and at present the best we can do is to attempt a phenomenological approach to take them into account , without having to deal with their detailed physical understanding .
it is however worth recalling that many attempts have been made to tackle the problem of injection at a more fundamental level ( malkov 1998 ; malkov & vlk 1995 ; malkov & vlk 1998 ) . here
, we consider the reasonable situation in which @xmath107 , where @xmath108 is the larmor radius of the particles in the downstream fluid that carry most of the thermal energy , namely those with momentum @xmath109 ( @xmath110 here is the momentum of the particles in the thermal peak of the maxwellian distribution in the downstream plasma , having temperature @xmath111 ) .
we stress here the important point that the temperature of the downstream gas ( and therefore @xmath112 ) is determined by the shock strength , which in the presence of accelerated particles , is affected by the pressure of the non - thermal component .
in particular , the higher the efficiency of the shock as a particle accelerator , the weaker its efficiency in terms of heating of the background plasma ( see section [ sec : heating ] ) . for collisionless shocks ,
it is not clear whether the downstream plasma can actually be thermalized and the distribution function be a maxwellian . on the other hand
, it is generally assumed that this is the case , so that in the following we consider the case in which the bulk of the background plasma is thermal and has a maxwellian spectrum at temperature @xmath113 given by the generalized rankine - hugoniot relations in the presence of accelerated particles ( see section [ sec : nonlin ] ) . for modified shocks ,
the points discussed above apply to the so - called subshock , where the injection of particles from the thermal pool is expected to take place .
we recall that for strongly modified shocks the subshock is weak , and rather inefficient in the heating of the background plasma .
[ fig : structure ] from eq .
[ eq : flux1 ] we get : @xmath114 where we assumed that the temperature downstream implies non - relativistic motion of the quasi - thermal particles ( @xmath115 ) . in eq .
[ eq : flux2 ] we write the minimum momentum in terms of a parameter @xmath116 , such that @xmath117 . with this formalism , the particles that can cross the shock surface are those that satisfy the condition : @xmath118 the parameter @xmath116 defines the thickness of the shock in units of the gyration radius of the bulk of the thermal particles . our recipe for injection
is pictorially illustrated in fig .
6 : thermal particles have a pathlength smaller than the shock thickness and can not cross the shock surface , being advected downstream before the crossing occurs . only particles with momentum sufficiently larger than the thermal momentum of the downstream particles can actually return upstream and be accelerated . in the following we will neglect the fluid speed @xmath102 compared with @xmath55 , which is a good approximation if the injected particles are sufficiently more energetic than the thermal particles .
this is done only to make the interpretation of the result simpler , but there is no technical difficulty in keeping the dependence of the results on @xmath102 .
we introduce an effective injection momentum @xmath119 defined by the equation : @xmath120 which in terms of dimensionless quantities , with @xmath121 reads : @xmath122 it is easy to show that @xmath123 for @xmath124 ( half a larmor rotation of the particles with momentum @xmath125 inside the thickness of the shock ) and @xmath126 for @xmath127 ( one full larmor rotation of the particles with momentum @xmath125 inside the _ thickness _ of the shock .
the fraction of particles at momentum @xmath128 times larger than the thermal one is @xmath129 for @xmath130 and @xmath131 for @xmath132 .
the actual values of @xmath128 are expected to be somewhat higher if the effect of advection with the downstream fluid is not neglected .
the sharp decrease in the fraction of _ leaking _ particles that may take part in the acceleration process is due to the exponential behaviour of the maxwellian at large momenta . although the fraction of particles in the maxwellian that get accelerated only depends on the parameter @xmath128 which in turn is expected to keep the information about the microscopic structure of the shock , the absolute number of and energy carried by these particles depend on the temperature of the downstream gas , which is an output of our calculation .
this simple argument serves as an explanation of the physical reason why there is a nonlinear reaction on the system due to injection . if the parameter @xmath128 is assumed to be determined by the microphysics of the shock , and if we adopt our simple recipe to describe such microphysics , then the shock thickness is easily estimated once the temperature of the downstream gas is known , and the latter can be calculated from the modified rankine - hugoniot relations .
the parameter @xmath30 in eq .
[ eq : inje1 ] is no longer a free parameter , being related in a unique way to the parameter @xmath128 and to the physical conditions at the subshock .
the condition that fixes @xmath30 is that the total number of particles in the non - thermal spectrum equals the number of particles in the maxwellian at momenta larger than @xmath23 . due to the very strong dependence of the spectrum on the momentum for both the maxwellian and the power law at low momenta , the condition described above is very close to require the continuity of the distribution function , namely that @xmath133 . in the following
we adopt this condition for the calculations .
this can be shown to imply the following expression for @xmath30 : @xmath134 we recall that the compression factor at the subshock , @xmath34 , approaches unity when the shock becomes cosmic ray dominated .
this makes evident how the backreaction discussed above works : when the shock becomes increasingly more modified , the efficiency @xmath30 tends to decrease , limiting the amount of energy that can be channelled in the non - thermal component .
although the recipe provided here is certainly far from representing the complexity of the reality of injection of particles from the thermal pool , it may be considered as a useful attempt to include the main physical aspects of this phenomenon .
in this section we describe the role played by the injection recipe discussed above for the appearance of multiple solutions .
it can be expected that the phenomenon is somewhat reduced because , as discussed in the previous section , the injection provides an efficient backreaction mechanism on the shock as a particle accelerator .
indeed we find that the appearance of multiple solutions is drastically reduced , and that the phenomenon still exists only in regions of the parameter space which are very narrow and of limited physical interest . in the quantitative calculations we use the value @xmath135 for the injection parameter , as suggested by the simple estimate in section [ sec : injectionrecipe ] and as suggested also in the numerical work of kang & jones ( 1995 ) .
the dependence of the effect on the value of @xmath128 is discussed below . in fig .
[ fig : single ] we illustrate the dramatic change in the physical picture by plotting @xmath68 as a function of @xmath33 for @xmath135 and adopting the same values for the parameters as those used in obtaining fig .
the efficiency @xmath30 is now calculated according with the recipe described in the previous section .
it can be seen very clearly that when the mach number of the shock is changed , there is a single solution ( compare with fig . 2 where multiple solutions where found for the same values of the parameters , but without thermal leakage ) .
[ fig : parspace ] the appearance of multiple solutions can be investigated in the whole parameter space , in order to define the regions where the phenomenon appears , when it does . in fig .
8 we highlight the regions where there are multiple solutions ( dark regions ) in a plane @xmath136 , for different values of the mach number of the shock . in most cases
the dark regions are very narrow and cover a region of values of @xmath128 which is rather high ( small efficiency ) . in fig .
[ fig : transition ] we plot the value of @xmath33 as a function of @xmath128 for @xmath137 , @xmath138 and @xmath139 from left to right .
the line is continuous when there are no multiple solutions and dashed when multiple solutions appear .
the dashed regions are , as stressed above , rather narrow .
for instance , for @xmath140 there are multiple solutions only for @xmath141 .
any small perturbation of the system that changes the values of @xmath128 at the percent level implies that the system shifts to one of the single solutions if it is sitting in the intermediate solution before the perturbation .
the sharp transition between the strongly modified solution and the quasi - linear solution when @xmath128 is increased suggests that the intermediate solution may be unstable , though a formal demonstration can not be provided here . in order to make sure that this is the case
, a careful analysis of the stability is required , and will be presented in a forthcoming publication ( blasi & vietri , in preparation ) . on the other hand ,
a previous study , carried out in the context of the two - fluid models , showed that when multiple solutions are present , the solution with intermediate efficiency is in fact unstable to corrugations of the shock surface ( mond and drury 1998 ) .
it is rather remarkable that the kinetic model of blasi ( 2002 , 2004 ) does not require explicitely the use of the equation for energy flux conservation .
however , once the solution of the kinetic problem has been found , the equation for conservation of the energy flux provides very useful information , as we show below .
the equation can be written in the following form : @xmath142 @xmath143 where @xmath144 is the flux of particles escaping at the maximum momentum from the upstream section of the fluid ( berezhko & ellison , 1999 ) .
notice that this term is usually neglected in the linear approach to particle acceleration at shock waves because the spectra are steep enough that , in most cases , we can neglect the flux of particles leaving the system at the maximum momentum .
the fact that particles leave the system make the upstream fluid behave as a radiative fluid , and makes it more compressible .
this is a crucial consequence of particle acceleration at modified shocks , and is shown here to be a natural consequence of energy conservation .
[ fig : escape ] in eq .
[ eq : energy ] we can divide all terms by @xmath145 and calculate the normalized escaping flux : @xmath146 @xmath147 from momentum conservation at the subshock we also have : @xmath148 so that the escaping flux only depends upon the _ environment _ parameters ( for instance the mach number at upstream infinity ) and the compression parameter @xmath34 which is part of the solution . note also that the adiabatic index for cosmic rays , @xmath149 , is here calculated self - consistently as : @xmath150 where @xmath151 is the energy density in the form of accelerated particles and @xmath152 is the kinetic energy of a particle with momentum @xmath20 .
it can be easily seen that @xmath153 when the energy budget is dominated by the particles with @xmath154 ( namely for strongly modified shocks ) and @xmath155 for weakly modified shocks . in eq .
[ eq : energy_norm ] the term @xmath156 is clearly the fraction of flux which is advected downstream with the fluid . in fig .
10 we plot the escaping flux ( @xmath157 ) , the advected flux ( @xmath158 ) and the sum of the two ( @xmath159 ) normalized to the incoming flux @xmath145 , as functions of the mach number at upstream infinity @xmath63 .
here we used @xmath160 , and @xmath135 , while the maximum momentum has been chosen as @xmath161 in the left panel and @xmath162 in the right panel .
several comments are in order : * at low mach numbers the escaping flux is inessential , as one would expect for a weakly modified shock .
we recall that the escaping flux is due to the particles with momentum @xmath56 leaving the system from upstream infinity . for a weakly modified shock at low mach number
the spectrum is steeper than @xmath163 , so that the energy carried by the highest energy particles is a small fraction of the total .
* at large mach numbers the shock becomes increasingly more cosmic ray dominated , and for the cases at hand the total efficiency gets very close to unity , meaning that the shock behaves as an extremely efficient accelerator . at mach numbers around 4 on the other hand the total efficiency is around @xmath164 for @xmath161 and @xmath165 for @xmath162 , dropping fast below mach number 4 .
clearly the efficiency would be higher in this region for lower values of the parameter @xmath128 .
* despite the fact that the total efficiency of the shock as a particle accelerator is close to unity at large mach numbers , the fraction of the incoming energy which is actually advected toward downstream infinity is only @xmath166 at @xmath167 for @xmath161 .
most of the enegy flux in this case is in fact in the form of energy escaping from upstream infinity at the highest momentum @xmath56 . for @xmath162
the normalized advected flux roughly saturates at @xmath168 and is comparable with the escape flux at the same mach number . for a distant observer these escaping particles would have a spectrum close to a delta function around @xmath56 .
energy conservation has the natural consequence that a smaller fraction of the kinetic energy of the fluid is converted into thermal energy of the downstream plasma in cosmic ray modified shocks , compared with the case of ordinary shocks .
the reduction of the heating at nonlinear shock waves is fully confirmed by our calculation in the context of the injection recipe introduced in section [ sec : injectionrecipe ] . in fig .
[ fig : jump ] we plot the temperature jump between downstream infinity ( at temperature @xmath111 ) and upstream infinity ( at temperature @xmath169 ) .
the thick solid line is the jump as predicted by the standard rankine - hugoniot relations without cosmic rays .
the other lines represent the temperature jump at cosmic ray modified shocks with @xmath170 ( thin solid line ) , @xmath171 ( dashed line ) , @xmath172 ( dotted line ) and @xmath173 ( dash - dotted line ) .
such a drastic reduction of the downstream temperature is expected to reflect directly in the thermal emission of the downstream gas in those environments in which collisions are relevant .
note that for strongly modified shocks the compression factor between upstream infinity and downstream are much larger than for ordinary shocks , so that the downstream turns out to be denser but colder than in the linear case .
the missing energy ends up in the form of accelerated particles .
the effect of suppression of the heating in cosmic ray modified shocks also appears in the spectra of the particles ( thermal plus non - thermal ) in the shock vicinity . in fig .
[ fig : spec ] we show these spectra ( including the maxwellian thermal bump ) for @xmath160 , @xmath135 and @xmath174 .
the vertical dashed line shows the position of the thermal peak as expected in the absence of accelerated particles .
in fact this position should depend on the mach number , but the dependence is very weak for large mach numbers .
the positions of the thermal peaks clearly show the effect of cooler downstream gases for modified shocks . at the same time
, the effect is accompanied by increasingly more modified spectra of accelerated particles , with most of the energy pushed toward the highest momenta .
the efficiency of the first order fermi acceleration at shock fronts depends in a crucial way upon details of the mechanism that determines the injection of a small fraction of the particles from the thermal pool to the _ acceleration box_. in reality the processes of formation of a collisionless shock wave , of plasma heating and particle acceleration are expected to be all parts of the same problem , though on different spatial scales .
we hide our lack of knowledge of the microphysics of the shock structure in a simple recipe for injection , in which the particles that take part in the acceleration process are those that have momentum larger by a factor @xmath128 than the momentum of the thermal particles in the downstream fluid .
this is motivated by the fact that for collisionless shocks the thickness of the shock surface is determined by the gyromotion of the bulk of the thermal particles .
we estimated that @xmath175 .
this recipe implies that the fraction of particles that get accelerated is rather small ( @xmath176 ) .
we implemented this recipe in a calculation of the non - linear reaction of cosmic rays on the shock structure proposed by blasi ( 2002 , 2004 ) . similarly to other models , also this approach shows the appearance of multiple solutions , for a wide choice of the parameters of the problem .
when the simple model of particle injection is used , this phenomenon is drastically reduced : the multiple solutions disappear for most of the parameter space , and when they appear they look as narrow strips in the parameter space , at the boundary between unmodified and modified shocks .
we argued that this result suggests that the narrow regions may signal the transition between two stable solutions , although this needs further confirmation through detailed analyses of the stability of the solutions .
this interpretation seems to be supported in part by the calculations of mond & drury ( 1998 ) , that showed that when three solutions are present , the intermediate one is indeed unstable for small corrugations of the shock front .
this calculations was however performed in the context of a two - fluid model , while an investigation of the stability for kinetic models is still lacking .
we find that the phenomenology of the particle acceleration at modified shocks is characterized by three main features : * the modification of the shock increases with the mach number of the fluid
. for low mach numbers the quasi - linear solution is recovered , but departures from it are evident already at relatively low mach numbers .
the modification of the spectra manifests itself with a hardening at high momenta and a softening at low momenta .
the @xmath177 shows a characteristic dip at intermediate momenta , typically around @xmath178 ( for very large values of @xmath56 , the dip can be found at even larger momenta , which is of interest for the acceleration of ultra - high energy cosmic rays ) .
* the total efficiency for particle acceleration saturates at large mach numbers at a number of order unity .
however , as shown in fig .
10 , the largest fraction of the energy is not advected downstream but rather escapes from upstream infinity at the maximum momentum .
this effect was also discussed in the context of the simple model by berezhko & ellison ( 1999 ) . *
the high efficiency for particle acceleration reflects in a reduced ability of cosmic ray modified shocks in the heating of the background plasma .
this effect is at the very basis of the backreaction introduced by the injection recipe on the acceleration process , and determines the saturation of the total efficiency for large mach numbers .
the heating suppression is shown in fig .
[ fig : jump ] and in fig .
[ fig : spec ] .
pb is grateful to e. amato , l.oc .
drury , d. ellison and m. vietri for many useful discussions .
the research of pb was partially funded though cofin-2004 at the arcetri astrophysical observatory .
sg gratefully acknowledges support from the alexander von humboldt foundation .
he also thanks stephane barland for useful discussions on nonlinear systems . | the dynamical reaction of the particles accelerated at a shock front by the first order fermi process can be determined within kinetic models that account for both the hydrodynamics of the shocked fluid and the transport of the accelerated particles .
these models predict the appearance of multiple solutions , all physically allowed .
we discuss here the role of injection in selecting the _ real _ solution , in the framework of a simple phenomenological recipe , which is a variation of what is sometimes referred to as _ thermal leakage_. in this context we show that multiple solutions basically disappear and when they are present they are limited to rather peculiar values of the parameters .
we also provide a quantitative calculation of the efficiency of particle acceleration at cosmic ray modified shocks and we identify the fraction of energy which is advected downstream and that of particles escaping the system from upstream infinity at the maximum momentum .
the consequences of efficient particle acceleration for shock heating are also discussed .
cosmic rays ; high energy ; origin ; acceleration |
in an early paper on the interaction between radiation and molecular dipole moments @xcite , dicke formulated a model which was later shown to exhibit a super - radiant phase transition@xcite .
the notion that such phase transitions should exist in condensed matter systems has been investigated in a series of papers by preparata and coworkers@xcite and others@xcite .
different workers have come to somewhat different conclusions concerning super - radiant phase transitions @xcite .
some doubt has been expressed@xcite concerning the physical laboratory reality of super - radiant phase transition .
the mathematical issues are as follows : ( i ) it appears,_at first glance _ , that quadratic terms ( in photon creation and annihilation operators ) enter into the model via quadratic terms in the vector potential @xmath0 .
( ii ) the quadratic terms in the `` corrected dicke model '' appear to destroy the super - radiant phase transition .
our purpose is to show that if the dipole - field interaction is treated in a _ gauge invariant manner_@xcite then the interaction is _ strictly linear _ in the electric field @xmath1 .
thus , quadratic terms are _ absent _ for purely electric dipole - photon interactions [ 28 ] .
these considerations render likely the physical reality of condensed matter super - radiant phase transitions .
to see what is involved , let us first consider a single molecule with an electric dipole moment @xmath2 given by @xmath3 where @xmath4 and @xmath5 denote respectively , the charge and position of the @xmath6 particle .
suppose that the dipole moment interacts with a _ classical _ electric field which is uniform in space , but not in time @xmath7 the molecule will have a time dependent hamiltonian of the form @xmath8 where @xmath9 is the kinetic energy of @xmath10 particle in the molecule and @xmath11 is the internal coulomb energy of the molecule .
employing the unitary operator @xmath12 in the time dependent canonical transformation @xmath13 yields @xmath14 for completeness of presentation , a derivation of eq.(5 ) is given in appendix a. note : for classical electric fields of the form in eq.(2 ) , the hamiltonians in eqs.(3 ) and ( 6 ) are _ rigorously _ equivalent .
furthermore , the hamiltonian in eq.(3 ) depends on linear and quadratic terms in the vector potential @xmath15 . on the other hand ,
the hamiltonian in eq.(6 ) depends only on the electric field @xmath16 and in a strictly _ gauge invariant _ and _ linear _ fashion . in the work which follows , both the charged particles and the electromagnetic field
will be treated using quantum mechanics . in sec.ii
we discuss a single molecule in the presence of the quantized electromagnetic field .
it is proved for the gauge invariant description of the dipole interactions that the coupling between the electromagnetic field and the molecule is linear in photon creation and annihilation operators the results for @xmath17 molecules are discussed in sec.iii .
the resulting hamiltonian has four terms : ( i ) a sum of single molecule hamiltonians , ( ii ) a sum of screened coulomb interaction potentials between neighboring molecules , ( iii ) the radiation field energy and ( most importantly ) ( iv ) a _ strictly linear coupling _ between the dipole moments and the electric field .
dicke - preparata models are defined and explored in sec.iv . in the concluding sec.v , the importance of the super - radiant phase transition associated with linear couplings will be discussed .
the errors by some previous workers who falsely `` voided '' the super - radiant phase transition will be discussed in detail .
consider , as in sec.i , a single molecule having a total number of @xmath18 charged particles with an electric dipole moment @xmath2 as in eq.(1 ) .
the dipole is considered to interact with a quantized electromagnetic field ( photons ) @xmath19 and/or @xmath20 which is again uniform in space .
the hamiltonian for the photons is given by @xmath21 where @xmath22 = \delta_{kk^\prime } .\ ] ] furthermore , for a `` quantization box '' of volume @xmath23 , @xmath24 and @xmath25 the matter hamiltonian for a single molecule in the electric dipole limit is then @xmath26 in total , the time independent hamiltonian for the molecule with dipole - quantized electromagnetic field interactions reads @xmath27 where @xmath28 is defined in eq.(7 ) , and @xmath29 is defined in eqs.(9 ) and ( 11 ) . in the quantum electrodynamic model , eq.(2 )
is replaced by the operator equation @xmath30\ ] ] leading to eq.(10 ) . employing the unitary operator @xmath31 in the time - independent canonical transformation @xmath32 yields @xmath33 where the `` self energy '' contribution to the molecule
is given by @xmath34 the derivation of eq.(16 ) is given in appendix b. introducing the boson operators @xmath35 and the hermitian conjugate @xmath36 yields our final hamiltonian @xmath37 @xmath38 where @xmath39 and @xmath40 = \delta_{kk^\prime } .\ ] ] the central result of this section is the following : * theorem : * for the electric dipole - electromagnetic field interaction , the resulting hamiltonian is that of free photons plus interaction terms linear in the photon creation and annihilation operators . _ quadratic interaction terms rigorously vanish . _ * proof : * see eqs.(18 ) and ( 19 ) , which were derived solely on the basis of the dipole interaction .
let us now consider the case of @xmath17 molecules , each of which interact with the quantized electromagnetic field via the electric dipole moment .
we consider the case of many molecules with electric dipole moments interacting with electromagnetic field modes whose minimum wavelength @xmath41 is large on the scale of the molecular size @xmath42 ; i.e. @xmath43 as shown in fig.1 below .
the hamiltonian for @xmath17 molecules plus the electromagnetic field reads @xmath44 the radiation hamiltonian in the coulomb gauge @xmath45 is given by @xmath46 the internal @xmath47 molecular hamiltonian is given by @xmath48 @xmath49 where @xmath50 is the central position of the @xmath51 molecule .
the internal hamiltonian of each molecule contains the internal coulomb potentials . in the last term on the right hand side of eq.(21 ) , @xmath52 describes the coulomb interactions between two different molecules .
the long ranged part of this potential is of the conventional dipole - dipole interaction form @xmath53 where @xmath54 .
the commutation relations of the quantum electrodynamic fields in the coulomb gauge are given by @xmath55= 4\pi i \hbar c { \bf \delta } ( { \bf r}-{\bf r}^\prime ) \ ] ] where the transverse delta function is defined as @xmath56 equivalently @xmath57 so that @xmath58 explicitly@xcite @xmath59 eqs.(25 ) and ( 30 ) imply @xmath60 the transverse distribution function @xmath61 removes the longitudinal part of a vector field .
for example , if @xmath62 represents polarization , i.e. the molecular dipole moment per unit volume , then @xmath63 represents the transverse part of the polarization .
now , consider the unitary operator @xmath64 eq.(26 ) , ( 33 ) and ( 34 ) imply @xmath65 so that the radiation hamiltonian of eq.(23 ) reads @xmath66 eqs.(17 ) , ( 23 ) , ( 31)-(33 ) and ( 36 ) imply that @xmath67 @xmath68 it is useful to define the `` screened '' ( or short ranged ) intermolecular coulomb potential with the long ranged dipole - dipole potential subtracted @xmath69 finally , the unitary transformation removes the vector potential from the internal degrees of freedom of a molecule as in the above sec.ii ; i.e. in the dipole interaction limit @xmath70 the `` grand finale '' of the algebraic ceremony of this section is the assertion that total transformed hamiltonian @xmath71 which for @xmath72 molecules takes the form @xmath73 @xmath74 where the internal molecular hamiltonian @xmath75 is independent of the transverse @xmath0 .
the central eq.(41 ) of this section follows from eqs.(17 ) , ( 24 ) , ( 37)-(39 ) and ( 40 ) .
the hamiltonian in eq.(41 ) has four parts : ( i ) a sum of intramolecular hamiltonians describing the internal kinetic and coulomb energies of each molecule , ( ii ) a sum of screened ( short ranged ) intermolecular coulomb potentials between neighboring molecules , ( iii ) the radiation field energy and ( iv ) a linear coupling between the dipole moment per unit volume and the electric field . within the dipole approximation , eq.(41 )
is exact but not yet amenable to rigorous mathematical solution . for the purpose of obtaining reasonable answers for ( say ) the free energy ,
further simplifications must be made .
for the dicke - preparata model , the approximations are as follows : ( i ) the sum of intramolecular hamiltonians is modeled by a sum of `` two - level molecules '' , each described by pauli matrices @xmath76 ; @xmath77 ( ii ) the sum of screened intermolecular coulomb potentials between neighboring molecules is neglected @xmath78 ( iii ) the radiation field energy is modeled by a single photon oscillator mode @xmath79 ( iv ) the linear coupling between the dipole moment per unit volume and the electric field is modeled by the _
resonant _ photon creation and annihilation terms @xmath80 where @xmath81 and @xmath82 is the excitation matrix element of the electric dipole operator of a single molecule . putting the model replacement eqs.(43)-(46 ) into eq.(41 ) yields the dicke - preparata model @xmath83 where @xmath84 @xmath85 and @xmath86 the dicke - preparata model may be written in terms of the oscillator coordinate @xmath87 and momentum @xmath88 using @xmath89 yielding @xmath90 where the vector @xmath91 is given by @xmath92 the free energy @xmath93 of the model is determined by @xmath94 here , the trace over the oscillator degree of freedom may be taken as being classical with virtually zero error in the limit @xmath95 ; i.e. @xmath96 since @xmath97 and since eq.(53 ) implies @xmath98 it follows from eqs.(52 ) , ( 53 ) , and ( 55)-(57 ) that @xmath99 where @xmath100 the oscillator energy per molecule is @xmath101 and the number of molecules per unit volume @xmath102 is @xmath103 in the thermodynamic limit of eqs.(58)-(61 ) , the minimum free energy principle is that @xmath104 where @xmath105 and @xmath106 let @xmath107 represent the mean oscillator energy per molecule as a function of the molecular density @xmath108 and the temperature @xmath109 .
if @xmath110 , then the system is in a super - radiant phase . if @xmath111 , then the system is in a normal ( incoherent radiation ) phase .
the free energy per molecule @xmath112 is at a minimum for @xmath113 .
setting @xmath114 yields the following implicit function equation for @xmath115 : @xmath116 the phase diagram in the @xmath117 plane follows from setting @xmath118 and solving for the critical temperature @xmath119 as a function of density .
the super - radiant phase exists for @xmath120 and the normal phase exists for @xmath121 .
the critical temperature @xmath122 where the critical density @xmath123 analytically , @xmath124 a plot of the phase diagram is shown below .
the doubts about the physical reality of the super - radiant phases were discussed in great detail by i. bialynicki - birula and k. rzazewski@xcite , who came to the remarkable ( and dubious ) conclusion that the electric dipole - photon interaction could be gauged away .
this incorrect conclusion was based on using a vector potential which is independent of both space and time .
it is no wonder that such a pure `` gauge field '' had zero effect on the thermodynamic equations of state . in reality
, the transverse electric field in the coulomb gauge should be calculated using either @xmath125 or @xmath126 , \ \
{ \rm ( schrodinger\ picture)}.\ ] ] the _ classical canonical transformation _
eq.(12 ) of i. bialynicki - birula and k. rzazewski@xcite ( in the dipole approximation ) should have been properly performed as in our eqs.(5 ) and ( 6 ) .
the last term on the right hand side of both our eqs.(5 ) and ( 6 ) were incorrectly omitted in eq.(12 ) of i. bialynicki - birula and k. rzazewski@xcite .
this incorrect classical canonical transformation eliminated the electric field @xmath127 and thereby the electric dipole - photon interaction .
a proper procedure is to employ a fully quantum electrodynamic framework to obtain a non - trivial electric field @xmath127 _ before _ going to the classical electrodynamic oscillator trace evaluation . from this rigorous schrdinger picture point of view ,
eq.(12 ) of i. bialynicki - birula and k. rzazewski@xcite is still _ incorrect _ because eqs.(b2-b3 ) in our appendix b eqs.(b2-b3 ) were not taken into account in that work@xcite . in any or all pictures
, the electric dipole - photon interaction _ can not _ be described by gauging away the electric field @xmath20 . when the hamiltonian is expressed in terms of @xmath20 and @xmath128 ,
it is then properly gauge invariant and non - trivial changes _ very surely _ appear in the thermal equations of state .
for example , the one - photon loop corrections to the equations of state have long been known to produce casimir - van der waals forces in condensed matter .
such forces can not possibly be gauged away and i. bialynicki - birula and k. rzazewski@xcite are thereby in error .
a further error is found in k. rzazewski , k. wodkiewicz and w. zakowicz@xcite , where it is asserted that the super - radiant phase transition is null and void due to quadratic coupling in the photon creation and annihilation operators . had the canonical transformation
been carried out correctly , as in the present work , then it would have been noted that the coupling between the electric dipole moment and the photons in the gauge invariant eq.(41 ) is _ strictly linear_. the super - radiant phase transition remains intact .
preparata has recently proposed several examples of super - radiant phases in a recent review@xcite .
water is perhaps the most interesting example@xcite . in future work
it would appear interesting to relate the super - radiant transition to an instability of the incoherent radiation phase which produces the conventional theoretical and experimental qed casimir effects .
qed effects at the one photon loop level have long proved their importance in the understanding of the long ranged intermolecular forces@xcite .
starting from the schrdinger equation @xmath129 one seeks a solution of the form @xmath130 where @xmath131 . putting eq.(a2 ) into eq.(a1 ) yields @xmath132 @xmath133 or equivalently @xmath134 where @xmath135 eq.(a5 ) is the required time dependent canonical transformation . from eqs.(1 ) and ( 14 ) it follows that @xmath136 eqs.(b1 ) and ( 11 ) then imply @xmath137 thus , the vector potential in the dipole interaction theory may be `` gauged away ''
. however , the dipole interaction returns to the full hamiltonian since eqs.(9 ) and ( 14 ) imply that @xmath138 and @xmath139 from eqs.(7 ) , ( b2 ) and ( b3 ) it follows that @xmath140 where the operator electric field @xmath127 is given in eq.(10 ) , and the molecular self energy @xmath141 is given in eq.(17 ) .
thus the same canonical transformation which gauges away the vector potential in the molecule hamiltonian , brings back the electric dipole interaction @xmath142 when applied to the radiation hamiltonian via eq.(b4 ) . finally , from eqs.(b2 ) and
( b4 ) it follows that @xmath143 @xmath144 and eqs.(16 ) and ( 18 ) then follows from eq.(b5 ) .
the proof of eq.(18 ) has been completed .
dicke , _ phys .
rev . _ * 93 * , 99 ( 1954 ) .
k. hepp and e.h .
lieb , _ phys .
* a8 * , 2517 ( 1973 ) . k. hepp and e.h .
lieb , _ ann .
phys . _ * 76 * , 360 ( 1973 ) .
e. del giudice , g. preparata and g. vitiello _ phys .
lett . _ * 61 * , 1085 ( 1988 ) .
g. preparata , `` quantum field theory of super radiance '' in _ problems of fundamental modern physics _ ,
r. cherubini , p. dalpiaz and b. minetti , world scientific , singapore ( 1990 ) .
g. preparata , _ phys .
_ * a38 * , 233 ( 1988 ) .
enz , _ helv .
. acta _ * 70 * , 141 ( 1997 ) .
s. takeno and m. nagashima , _ prog .
phys . _ * 57 * , 1507 ( 1977 ) .
y.k . wong and f.t .
hioe , _ phys .
* a7 * , 831 ( 1973 ) .
carmichel , c.w .
gardiner and d.f .
walls , _ phys .
_ * a46 * , 47 ( 1973 ) .
g. vertogem and a.s .
de vires , _ phys .
lett . _ * a48 * , 451 ( 1974 ) .
hioe , _ phys . rev . _
* a8 * , 1440 ( 1973 ) .
kudenko , a.p .
slivinsky and g.m .
zaslavsky , _ phys .
lett . _ * a50 * , 411 ( 1975 ) .
b.m . pimentel and a.h .
lett . _ * a53 * , 200 ( 1975 )
zaslavsky , y.a .
kudenko and a.p .
slivinsky _ sov . phys
. solid state _ * 16 * , 2197 ( 1975 ) .
r. gilmore and c.m .
bowden , _ phys .
rev . _ * a13 * , 1898 ( 1976 ) .
r. gilmore and c.m .
* 17 * , 1617 ( 1976 ) .
a.a.s . brito and w.f .
wreszinski , _ helv .
acta _ * 50 * , 373 ( 1977 ) .
a. klemm and v.a .
zagrebnov , _ physica _ * a36 * , 400 ( 1977 ) . k. rzazewski , k. wodkiewicz and w. zakowicz , _ phys .
lett . _ * 35 * , 432 ( 1975 ) .
k. rzazewski and k. wodkiewicz , _ phys .
rev . _ * a13 * , 1967 ( 1976 ) .
k. rzazewski , k. wodkiewicz and w. zakowicz , _ phys .
* a58 * , 211 ( 1976 ) .
i. bialynicki - birula and k. rzazewski , _ phys .
_ * a19 * , 301 ( 1979 ) .
k. gawedzki and k. rzazewski , _ phys .
rev . _ * a23 * , 2134 ( 1981 ) .
woolley , _ j. phys
. lett . _ * a9 * , l15 ( 1976 ) .
woolley , _ adv .
phys . _ * 33 * , 153 ( 1975 ) .
power and s. zienau , _ phil .
_ * a251 * , 427 ( 1959 ) .
g. preparata , e. del giudice and r. mele , _ mod .
phys . lett . _ * b7 * , 1851 ( 1993 ) .
d.p . craig and t. thirunamachandran , _ `` molecular quantum electrodynamics '' _ , p 56 , dover publications , new york ( 1998 ) .
g. preparata , _
`` qed coherence in matter '' _ , world scientific , singapore ( 1995 ) . e. del giudice and g. preparata _ `` a new qed picture of water '' _ in _ `` macroscopic quantum coherence '' _ edited by e. sassaroli , y. srivastava , j. swain and a. widom , world scientific , singapore ( 1998 ) . | in a gauge invariant formulation of the molecular electric dipole - photon interaction , the rigorous coupling is strictly linear in the photon creation and photon annihilation operators .
the linear coupling allows for a super - radiant phase transition as in the hepp - lieb formulation .
a previous notion of a quadratic - coupling `` no - go theorem '' for super - radiance is incorrect . also incorrect
is a previous assertion that the dipole - photon coupling has absolutely no effect on the thermal equations of state .
these dubious assertions were based on incorrect canonical transformations which eliminated the electric field ( and thereby eliminated the dipole - photon interaction ) which is neither mathematically nor physically consistent .
the correct form of the canonical transformations are given in this work which allows for the physical reality of super - radiant condensed matter phases . |
gamma - ray burst ( grb ) spectra often extend to very high energies with no indication of attenuation by photon - photon interactions .
this implies substantial relativistic bulk motion of the radiating material with lorentz factors in the range of @xmath13 to @xmath14 . at cosmological distances ,
grbs require an energy reservoir on the order of @xmath15 erg .
the likely sources of such a reservoir would be the rest mass of a compact object released during a merger ( e.g. , either neutron star
neutron star or neutron star black hole ) .
however , most detailed calculations of mergers occur on time scales of less than 1 s , much less than the observed durations of grbs ( often @xmath13 s and sometimes @xmath14 s ) .
two classes of models have arisen that explain various ( but not all ) aspects of the observations . in the `` external '' shock models ( @xcite ) ,
the release of energy during the merger is very quick , and a relativistic shell forms that expands outward for a long period of time ( @xmath16 to @xmath17 s ) . at some point ,
interactions with the external medium ( hence the name ) cause the energy of the bulk motion to be converted to gamma - rays . although the shell might produce gamma - rays for a long period of time , the shell keeps up with the photons such that they arrive at a detector over a relatively short period of time . if the shell has a velocity , @xmath18 , with a corresponding bulk lorentz factor , @xmath19 , then photons emitted over a period @xmath20 arrive at a detector over a much shorter period , @xmath21 .
although this model is consistent with the short energy release expected for a merger and the observed long time scale of grbs , we have argued that it can not explain the long complex time histories except under extreme conditions ( @xcite ) . in particular , we argue from kinematic considerations that the long gaps often seen in grbs are inconsistent with local spherical symmetry , that the rapid time variability implies that only a small fraction of the shell becomes active , and that the observed average envelope of emission is inconsistent with that expected from a relativistic shell .
these arguments are contained in @xcite , and @xcite .
in addition , @xcite , @xcite , and @xcite argue from hydrodynamic considerations that the external shock model can not provide the energy or timescales observed in grbs .
the alternative theory is that a central site releases energy in the form of a wind or multiple shells over a period of time commensurate with the observed duration of the grb ( @xcite ) .
the gamma - rays are produced by the internal interactions within the wind ; hence these scenarios are often referred to as internal shock models .
these models have two weaknesses : first , there is a concern that internal shocks are rather inefficient ( although , see @xcite ) , and second , the long , complex time history of a grb must be postulated at the central site . on the other hand , the free parameters associated with the multiple shells
can probably explain any observed time history ( @xcite ) .
the need to explain long time histories was further complicated by the discovery of the x - ray afterglows lasting hours ( @xcite ) , the optical afterglows lasting weeks to months ( @xcite ) , and the radio afterglows lasting many months ( @xcite ) .
these extended time scales appear too long to have been produced by a lingering central site made by a merger .
in addition , the observed power law decay is expected from many external shock models ( @xcite ) .
@xcite suggested that the initial gamma - ray phase is due to internal shocks from a relativistic wind ( or multiple shells ) that merge into a single relativistic shell which then produces the afterglows in a manner similar to the external shock models .
this model avoids the difficulty of explaining the rapid variability of the gamma - ray phase with a single shell while retaining the long time scale capability of the single shell for the afterglows .
the purpose of this paper is to demonstrate that the external shock model can not utilize the full surface of shells .
fundamentally , it arises because of the relatively rapid time variability of the gamma - ray phase .
@xcite use this as an argument that external shock models would require too much energy .
however , this should not be considered a strong argument that the grb phase is not external shocks since there are several scenarios that can utilize only a portion of the shell s surface .
we also analyze the time variability recently observed by asca in the x - ray afterglow of gb970828 ( @xcite ) and show that its time variability implies low surface utilization in the x - ray afterglow phase as well .
to understand the problem caused by rapid variability , one must emphasize the difference between arrival time at the detector ( which we denote with @xmath8 ) and coordinate time ( equivalent to laboratory time , the detector rest frame time , _ and _ the rest frame of the central explosion , denoted by @xmath20 ) .
coordinate time is measured by clocks placed at all locations within the frame and can measure when the photons were produced .
in contrast , a detector ( such as batse ) measures when the photons arrive at a single location .
these two times are related as @xmath22 where @xmath23 is the angle between the direction of motion of the emitting material and the direction to the detector .
the large bulk lorentz factor means that the shell is almost certainly moving directly at the observer s detector ( i.e. , we must be looking at the shell head - on rather than looking at the sides of a jet , so @xmath23 is small . when @xmath23 is 0 , @xmath24 .
in contrast , the lorentz transformation between the time in the rest frame of the detector ( @xmath20 ) and time in a frame moving with the shell ( @xmath25 ) , is @xmath26 .
consider the early behavior of the shell .
assume that clocks are placed along the path of the shell in the detector s rest frame .
these clocks report coordinate time , that is , the time the shell passes the location of the clocks . consider an extreme case such as @xmath27 .
clocks placed with a spacing of @xmath28 cm will see the shell pass their locations at times separated by @xmath29 s. because the shell is viewed by the detector head - on , its movement causes the shell to keep up with the photons it produces such that photons that were separated by @xmath29 in the detector s rest frame , arrive at the detector separated by a much shorter interval @xmath30 .
if the shell expands in the detector rest frame for @xmath17 seconds , the photons all arrive at the detector within 5 s. note that this contraction of the photon arrival time is not the result of a lorentz transformation .
rather , it reflects that a detector at a single location measures when photons arrive , not necessarily the relative time they were emitted . in the early phase of the shell s expansion , @xmath3 is effectively constant ( @xmath31 ) such that @xmath32 .
eventually , the shell begins to deaccelerate as it sweeps up the interstellar medium ( ism ) . without detailed knowledge of the physical process that generates the gamma rays ,
it is not clear if the expansion is adiabatic or dominated by radiation losses .
however , both solutions lead to a power law decay ( @xcite ) , and we will show that our conclusions only depend on the power law nature of the decay and not the exact index of the power law . for the sake of an example , we will assume that the expansion is adiabatic .
the energy required to sweep up an ism mass of @xmath33 is @xmath34 ( @xcite ) .
let @xmath35 be the initial energy deposited into the shell such that @xmath36 , where @xmath37 is the original mass of the shell .
if @xmath38 is the density of the ism , one can solve for the radius of the shell as a function of the coordinate time by solving @xmath39 here , is the angular size of the ejecta to account for the situation where the expansion is mechanically confined to a jet .
the arrival time of the photons as a function of the coordinate time that the photons were emitted can be found from @xmath40 and @xmath41 .
figure 1 shows @xmath20 as a function of @xmath8 for a typical situation ( @xmath42 erg , @xmath43 atom @xmath44 ) .
it is unclear what fraction of the shell s energy is converted to gamma - rays .
observations imply that the energy in gamma - rays _ if the explosion is isotropic _ is on the order of @xmath15 erg . if is much less than @xmath45 , much less energy is required ( implying many more unseen events ) .
the observations only limit the energy per @xmath45 steradian to be @xmath46 erg .
we scale our solutions to @xmath42 erg which assumes that @xmath47 1% of the energy is converted to gamma - rays .
there are three phases of the @xmath48 relationship : the initial constant-@xmath3 phase , a phase where @xmath49 , and a final phase when the shell becomes nonrelativistic where @xmath20 is @xmath8 plus a constant . here
, @xmath50 is @xmath35 in units of @xmath51 erg and @xmath52 is the ism density in units of 1 atom @xmath44 .
the second and third phase is only weakly dependent on the initial conditions .
neither depends on the initial @xmath53 , and they depend on the other parameters to the 1/4 power . thus , except for the duration of the first phase , different bursts ought to have similar @xmath48 relationships . if one assumes the extreme nonadiabatic case that the momentum of the shell is conserved rather than the energy , then the 1/4 power becomes 1/7 .
many other indexes as possible ( see @xcite ) .
although the exact nature of the conversion of bulk motion to gamma - rays is unclear , we argue that the gamma - ray phase must be contained in the first phase where @xmath54 .
we note that many grbs show a similar time of arrival structure ( e.g. , observed subpeak widths ) at the end of the burst and at the beginning of the burst . in the first phase , the observed arrival time structure ( @xmath7 )
is related to the emission time as @xmath55 such that a constant subpeak width can occur if the emission time is constant in coordinate time . in the second phase , @xmath56 such that constant subpeak width can only occur if the emission time ( @xmath57 ) scales as @xmath58 .
this is consistent with our scaling of @xmath42 erg , which allows the first phase to last 200 s ( see figure 1 ) .
the afterglows probably occur in the @xmath59 phase where @xmath3 is more modest .
as one can see from figure 1 , our view of grbs becomes quite distorted because we only measure arrival time at the earth rather than the time of emission in the earth s rest frame ( remember , no lorentz transformations are involved ) .
it is often said that grbs are quite quick , say 30 s. actually , in the external shock models , they occur longer by @xmath60 or , perhaps , up to @xmath61 s if @xmath3 is @xmath14 .
the afterglow period seems to cover a very large dynamic range from @xmath14 s to @xmath17 s. actually , because @xmath59 , the dynamic range is only a factor of 10 .
in the external shock model , there is a single shell , and the rapid variability we observe can only be accounted for by a nonuniform production of gamma - rays by the shell .
consider batse bursts 1883 and 2993 shown in figure 2 .
they have very similar overall envelopes of emission and similar intensities . burst
2993 , however , has stronger deviations from the envelope in the form of multiple peaks .
presumably this structure arises because various patches ( or emitting `` entities '' ) on the shell randomly become active . in the case of burst 1883 , so many entities become simultaneously active that the overall envelope appears quite smooth , whereas in the case of burst 2993 , fewer entities become active so random fluctuations in the number of simultaneously active entities cause the peak structure to be spiky . in this view , each observed peak in not necessarily caused by a single entity , but the peak structure is caused by random variations in the number of active entities ( see @xcite ) . in @xcite and @xcite , we showed that the expected envelope of the grb phase has a `` fred '' like shape ( fast rise , exponential decay although the decays are often just slower , not necessarily exponential ) .
the expected envelope is a power law , @xmath62 , where @xmath63 is the spectral number index ( typically @xmath64 ) .
( note that this power law refers to the grb phase in constrast to the well observed power law decay found in the afterglows . ) in @xcite , we showed that the volume that contributes to a set period of time , @xmath7 , is constant .
the @xmath65 envelope comes from the beaming .
thus , if the entities form randomly , one can expect that the rate of entities causing peaks is effectively constant during the decay phase .
indeed , the presence of long gaps in grbs is one of our key arguments that the grb phase is not due to a simple single shell .
gaps require large causally disconnected regions to coordinate their lack of emission ( see @xcite and @xcite ) .
we define the `` surface filling factor '' as the fraction of the shell s surface that becomes active .
let @xmath66 be the area of an entity and @xmath67 be the number of entities that ( randomly ) become active during the interval @xmath68 . if @xmath69 is the area of the shell that can contribute during @xmath68 , then the surface filling factor is @xmath70 where @xmath71 is the fraction of the visible area of the shell ( @xmath72 ) that contributes during the interval @xmath68 .
the number of entities can be determined from the observed fluctuations in the time history .
the overall envelope is due to the beaming , so the residue variations about the overall envelope are due to a combination of the poisson variations in the number of entities and the poisson variations associated with the count statistics .
assume for the moment that the contribution due to the count statistics are small such that the variations come from the poisson variations in the number of contributing entities .
we first remove the envelope by fitting a polynomial function to it .
the observations are divided by the polynomial function , so the result is a flat envelope with variations due to the number of entities that , on average , are active simultaneously .
the rate of occurrence of the entities , @xmath73 , can be found directly from the variations because the observed mean level is @xmath74 where @xmath75 is some constant and the variance is @xmath76 .
we define @xmath77 to be the observed mean level of the flattened envelope and @xmath78 to be the root mean square of the flattened envelope . in the case of no contribution from the counting statistics ,
@xmath73 would be @xmath79 . here
, we have implicitly assumed that all entities are identical .
this is supported by the fact that peaks within grbs usually are similar to each other .
the actual root mean square of the flattened envelope is a combination of the variance due to the entities and the variance due to the counting statistics , @xmath80 .
we assume they add in quadrature , that is , @xmath81 .
we estimate @xmath82 to be root mean square of many monte carlo realizations of the flattened envelope .
the rate of occurrence of entities is @xmath83 this rate is the number of events per the time scale of the entities .
thus , the total number of entities that occur within a period @xmath68 is @xmath84 where @xmath85 is the time scale for a single entity .
we have consider several scenarios that relate the size of an entity causing a peak to an observed peak duration , @xmath85 ( see table 1 in @xcite ) .
here , we consider the two most likely scenarios for the formation of a peak : regions that grow and regions formed by the interaction with the ism .
consider a gamma - ray producing region that grows at the speed of sound , @xmath86 , for a period @xmath87 in the rest frame of the shell where presumably the region is symmetric .
this growth might be associated with a developing shock .
let @xmath88 be the radius of the region in the rest frame along the direction of the motion and @xmath89 be the radius in the perpendicular direction such that @xmath90 the sizes in the rest frame of the detector are related to the sizes in the rest frame of the shell as : @xmath91 and @xmath92 .
combining the effects of the movement of the shell during the growth with the maximum size that the entity can grow in time @xmath57 , we find that the duration in arrival time is ( see table 1 of @xcite ) : @xmath93^{1/2}~. \eqno(\neweq)\ ] ] the size of the emitting entity is @xmath94}~~. \eqno(\neweq)\ ] ] the alternative cause of a peak resulting from the shell is that the shell interacts with an ambient object such as an ism cloud .
presumably , the object is symmetric such that @xmath95 .
( we use lower case @xmath96 for an object that grows in a shell and upper case @xmath97 for an ambient object . )
we assume the case of a `` collapsible '' object ( such as an ism cloud , cf .
the contribution to the peak duration from time the shell takes to move through the cloud ( i.e. , @xmath98 $ ] ) is negligible compared to the time the shell takes to engage the perpendicular size of the object .
this engagement time is caused by the curvature of the shell . at an angle @xmath23 from the line of sight , the time to engage the object is @xmath99 . note
that table 2 of @xcite was incorrect for @xmath100 ; see @xcite .
at a typical angle of @xmath101 , @xmath102 and @xmath103 thus , both the case of shocks that grow from a seed and the case of shells running into ambient objects are similar . if @xmath104 , then @xmath66 from equation ( ) is 16/13 times larger than from equation ( ) .
these two scenarios only differ by a constant the order of unity .
the final ingredient for the calculation of the surface filling factor is the area of the shell visible to the observer : @xmath105 where @xmath106 is either the angular width of the shell [ i.e. , @xmath107 ) ] or it is @xmath108 , whichever is smaller . before the shell starts to deaccelerate , presumably @xmath106 is larger than @xmath109 . using @xmath110 ,
the above was derived for the adiabatic case .
more generally , equation ( b ) becomes @xmath119 where @xmath120 is 1 for the adiabatic case and 0 for the radiative ; @xmath121 is a parameter that varies from 0 to 1 depending on whether most of the energy or most of the mass is concentrated at low @xmath3 s ; and @xmath122 is a parameter that varies from 0 to 1 depending on whether the emitting electrons are radiatively efficient or adiabatic ( see @xcite ) .
more involved power laws are possible when one includes variations in the ambient material ( see @xcite ) .
the surface filling factor from equation ( ) depends on @xmath123 . for a power law relationship between @xmath20 and @xmath8 , @xmath124 . assuming that each entity is similar in the rest frame , different deaccelerations ( i.e. , various values of @xmath120 , @xmath122 , @xmath121 ) only modify the equation for @xmath5 by a numerical factor on the order of @xmath125 .
we conclude that surface filling factors found in the next section are roughly what one would find for any deacceleration scenario as long as @xmath126 .
the consistency of the peak duration throughout the gamma - ray phase implies that @xmath3 is rather constant . in this case , the surface filling factor is a function of @xmath71 ( from eq . [ ] ) , and the direct observables ( @xmath127 , and @xmath8 ) .
the constant @xmath116 is either 16 or 13 depending on whether the source of the entity is due to ambient objects or shocks that grow , respectively .
if ambient objects were involved , we would expect to see peaks with a wide range of durations , perhaps a power law distribution .
in fact , most grb peaks are about the same size ( @xcite ) , so we will use @xmath128 for the gamma - ray phase . to estimate @xmath71 , one needs to estimate the fraction of the visible shell ( i.e. , @xmath72 , the area within a beaming angle of @xmath2 ) that contributes to the signal during time @xmath8 .
unfortunately , we do not know how , in general , to map the amount of active area of a shell to the observed time history . only in the particular case of a fred - like burst
can we reliably estimate @xmath71 . for a fred - like burst
, the curvature of the shell delays the photons from the region of the shell at angle @xmath2 by @xmath129 where @xmath129 is the time that the shell expands before producing gamma - rays . from @xcite and @xcite
, one can find @xmath129 from the shape of the profile .
given that the profile shape is @xmath130 , @xmath129 is @xmath131 where @xmath132 is the duration of the burst that contains 50% of the event counts ( see batse catalog , @xcite ) .
here , @xmath63 is the spectral number index , typically @xmath64 .
thus , if one uses @xmath133 , then the profile is from the region of the shell out to an angle of @xmath2 , and @xmath134 . for long complex bursts
, we can not easily detect a fred - like shape . in lieu of an exact estimate
, we will use the rule found for fred - like bursts : @xmath135 . in table 1 , we list 6 fred - like bursts and 46 long complex bursts that were bright and had sufficient time structure for estimating parameters . the first column gives the batse trigger number .
the second column gives @xmath73 as found from equation ( ) .
the third column gives an estimate of @xmath85 , the temporal width of a single entity .
in most cases , this was estimated from the width of the autocorrelation function . for four of the fred - like bursts ,
the profiles were rather smooth ( presumably due to a larger number of entities being simultaneously active ) so we used a typical value for @xmath85 ( i.e. , 0.5 s ) .
the 4-th column is @xmath132 and the next two columns are , respectively , the surface filling factor and its propagated error based on equation ( ) . in figure 3
we show the distribution of surface filling factors as a function of burst duration @xmath132 .
the solid squares are the fred - like bursts , and the open squares are the long complex bursts .
although some of the smooth fred - like bursts can have surface filling factors near unity , most bursts have values on the order of @xmath136 . in figure 4 , we show the time history of burst gb970828 observed by the asca satellite , from 28.9 to 48.9 hours after the gamma - ray phase ( cf .
@xcite ) . at about 36.1 hours
, there was a flare that lasted about 2 hours .
@xcite also reported variations from a simple power law decay in the x - ray afterglow of gr970508 .
these two examples of deviations from a power law decay are quite different . in the case of the @xcite observation ,
the flare is relatively narrow ( i.e. , @xmath137 is small ) . in the case of the @xcite observation
, the deviation could be a step function increase in the intensity ( i.e. , @xmath138 , although there is a data gap from 22 to 65 hr after the grb that obscures the nature of the deviation ) .
we will emphasize the @xcite observation because the narrow nature of the flare gives much stricter limits on the relativistic shell ( see 5 ) . except for earth occultations ,
asca continuously observed the afterglow from 28.9 to 48.9 hr after the grb .
the observations have been prorated into samples of equal duration in figure 4 .
consider how case c in equation ( ) applies to gb970828 .
we must estimate the fraction of the shell that contributes to the 20 hours of observations , @xmath71 . we can estimate a lower limit on @xmath71 , and therefore , an upper limit on @xmath5 . in case c
we assume that the whole shell was active , but only a fraction ( @xmath71 ) of it contributed during the asca observation .
the narrowest annulus that contributes during any particular observation is one with a width equal to the size of a single entity . at a typical angle of @xmath2
, @xmath71 can be as small as @xmath139 at smaller angles , @xmath71 would be smaller , but that is counteracted by the fact that the peak in the asca observation is clearly much smaller than the whole observation , such that the annulus responsible for the observation must be larger than we used in equation ( ) .
we only have observations covering a small range of time , so we can not use equations ( ) and ( ) to estimate @xmath67 .
however , there are probably only a few active entities because we see only one peak in figure 4 . hence
, we will use @xmath140 as the maximum number of entities involved in the asca flare .
thus , using case c , @xmath5 is , at most , @xmath141 , a value very similar to what is obtained in the gamma - ray phase . in figure 3
, the asca flare for case c is represented by the solid triangle . in the situation of case
b , we do not know or @xmath71
. however , the smallest @xmath71 when the shell s angular size is limited to is @xmath142 such that @xmath143^{-1/4 } { \gamma \delta t_p \over t^{1/4 } } = { 1.6 \times
10^{-5}n_n \over \sqrt{{\rm d}\omega } } \bigg[\initialcon\bigg]^{-1/8 } { \delta t_p \over t^{5/8}}~~. \eqno(\neweq)\ ] ] using , @xmath144 and @xmath145 , the maximum @xmath5 under case b is @xmath146 . in summary , @xmath137 in the asca flare
is @xmath147 which implies a filling factor of @xmath9 .
the x - ray afterglow suffers from the same time history problems that the gamma - ray phase has : peaks do not have the value of @xmath137 expected from a shell whose surface quasi - uniformly emits photons .
the rapid variations in grb time histories imply emitting entities the size of @xmath148 . assuming a _ single _ expanding shell , these entities must form on a much larger surface , @xmath149 .
we have defined @xmath5 ( given by eq . [ ] ) to be the fraction of the surface of the shell which becomes active , that is , generates emitting entities . a crude estimate of @xmath5 is @xmath150 . because both the classic gamma ray burst phase and the recently discovered x - ray afterglow phase can have flares with @xmath151
, both time periods can suffer from an inefficient utilization of a single relativistic shell . in the case of the gamma - ray phase
, there is a simple explanation : the gamma - ray source is not a single relativistic shell , but rather , reflects activity at some central site that produces @xmath0 erg s@xmath1 for up to several hundred seconds .
other observations , particularly gaps seen in the gamma - ray time histories ( @xcite ) , also argue for a central engine .
the x - ray afterglow , however , is widely interpreted as being a single relativistic shell ( @xcite ) .
therefore , we will concentrate our discussion on how to accommodate flares during the x - ray afterglow .
these arguments could also be applied to the gamma - ray phase .
of the two examples of deviations from a power law in the x - ray afterglow , the @xcite event ( gb970828 ) is much more restrictive than the @xcite event ( gb970508 ) .
gb970828 has a narrow flare whereas gb970508 appears to be roughly a step function .
a step function is similar to a fred - like deviation which is approximately what one would _ expect _
if most of the observable surface of a relativistic shell interacts with a large ism cloud ( i.e. , @xmath137 should be @xmath152 , @xcite ) .
thus , gb970828 might be an afterglow with a low surface filling factor whereas gb970508 has a high surface filling factor .
a common misconception is that one can just use an ism cloud that covers most of the shell s surface and , therefore , make a flare that substantially changes the emission .
this does not work because the curvature of the expanding shell prevents the shell from engaging the cloud instantaneously .
rather , the portion of the shell at @xmath153 requires a time @xmath154 longer to reach the cloud . even if the cloud happens to have a concave shape such that the shell reaches the cloud simultaneously over a wide range of angles , the resulting photons at @xmath101 must travel further to the detector resulting in emission that is delayed by @xmath155 .
indeed , the delay of the photons due to the curvature is identical to the scenario of a shell that expands without producing photons for a long time and then emits only over a small range of times ( the @xmath156 case in @xcite ) .
the result is a fred - like flare ( like gb970508 ) whose duration , @xmath7 , is about @xmath8 , rather than a small fraction of @xmath8 as indicated by the asca observations for gb970828 .
the shell will have well - defined curvature because all points on it are moving at nearly the same speed , that is , the speed of light . in any case , there is no reason to believe that variations in the ambient material would cause the shell to develop into a plane wave oriented towards the observer such that the photons produced by an interaction with an ism cloud or a shock would arrive as a short flare . _
only the instantaneous interaction between two plain parallel surfaces oriented perpendicularly to our line of sight can produce a short peak from large surfaces .
_ we see six possible explanations for flares with small @xmath137 that produce large changes in the observed intensity .
these explanations are explained below and represented in figure 5 .
a key issue is the size of the corresponding energy reservoir required in each case . to intercompare these energy requirements
, we use that a typical peak corresponds to an isotropic luminosity , @xmath157 , equal to @xmath158 erg s@xmath1 .
the total energy reservior is required to have @xmath159 where @xmath160 is the average duration of the event ( say , 30 s ) and @xmath161 is the fraction of the total energy that is converted into gamma - rays ( say @xmath162 ) . if the emission is isotropic , then the total energy is about @xmath163 erg .
since we can not observe the energy released at angles much greater than @xmath164 , any model can reduce the required energy reservior to @xmath165 . although the energy requirement can be made small , a single shell that emits nearly uniformly over ( see fig .
5a ) is not consistent with the observed time structure since a uniform shell always results in a fred - like shape .
the following scenarios can produce bursts with peaks with small @xmath137 : 1 .
the simplest way to have short powerful flares is to invoke a central engine . in a central engine , there are multiple releases of energy at the central site over a time scale commensurate with the observed duration of the grb .
each release makes a peak in the time history .
indeed , small @xmath137 and gaps are the key kinematic reasons why the grb phase is likely to be a central engine ( @xcite ) .
it is possible for a large fraction of the surface of the shell to be active , and the flare region might still be small , because @xmath161 or the energy content of the shell varies strongly as a function of position on the shell .
that is , the photons @xmath166 s@xmath1 at the flare site must be larger than the rest of the shell by a factor of @xmath167 ( see fig .
5b ) . for example
, some photon production rates scale as @xmath168 , so a single region with a @xmath3 that is 10 times larger than other regions could produce emission that doubles the overall count rate
. this would give the appearance that @xmath5 is @xmath169 . if the energy content of the shell is constant as a function of angle and @xmath161 varies
, then the required total energy is about @xmath170 .
if @xmath161 is constant , and the energy content of the shell varies , then the total required energy is @xmath165 .
the third explanation is that the local spherical symmetry of the shell is broken into many emitting regions on a scale much smaller than @xmath2 where only a fraction @xmath5 of the surface converts its energy into photons .
the regions are small enough ( @xmath171 ) to make peaks with small @xmath137 ( see fig .
5c ) . here
it is assumed that all regions of the shell has equal energy content but much of it is wasted because most regions do not produce gamma - rays .
thus , this scenario requires an energy reservoir of @xmath170 , that is , @xmath172 _ more _ energy than a simple shell . if is large , this scenario could require @xmath173 to be as large as @xmath174 erg .
@xcite invoke this explanation and the resulting need for more energy to argue against a single relativistic shell .
that arguement should not be considered particularly strong since we have several scenarios that do not waste the energy and/or could be small .
the fourth explanation is that the local spherical symmetry is broken because the outflow of material is in narrow fingers that occupy only a fraction @xmath5 of the shell s surface ( see fig .
this is basically the same geometry as the previous case except those regions that do not emit gamma - rays do not have energy content .
since no energy is wasted in non - emitting regions , this scenario requires an energy reservoir of only @xmath165 , that is , about what one would expect from a simple shell .
the fifth explanation is that is very small , so small that the emission lasts only @xmath85 even if the entire region is gamma - ray active .
this requires to be @xmath175 .
the requried total energy is correspondingly small : @xmath176 erg .
this explanation has one narrow ejecta that needs to go through a gamma - ray emitting phase for each peak in the grb time history . in the above explanation ( number 4 ) , there are many narrow ejecta , so each one only has to become gamma - ray active only a few times to make a grb time history with hundreds of peaks .
the last explanation attributes the duration of the grb to the spreading out in time of the emission because of different conditions as a function of @xmath23 ( see fig .
this is what we called a `` thick shell with substructure '' in @xcite .
it would have been better to call it a `` thick emitting region with substructure '' since the actual emitting shell must always be thin ( cf .
effectively , @xmath5 is unity but it appears to be small because the assumptions involved in equation ( ) are not valid .
let @xmath177 be the time that the @xmath178-th section of a shell at @xmath179 expands before becoming active .
let @xmath180 be the speed of the material at @xmath179 .
the relative time that the emission from two different regions ( @xmath178 and @xmath181 ) arrive at our detector is @xmath182 explanations 3 and 4 , above , assume that all @xmath177 s and @xmath180 s are equal such that we can estimate the size of the shell from it s fred - like appearance and peaks arrive at different times because they come from different @xmath179 s .
of course , most bursts are not fred - like . in this explanation ,
all @xmath179 regions become active ( so @xmath5 = 1 ) , but the peaks that would have added up to make the fred - like envelope are scattered in time because of the different @xmath177 s and @xmath180 s . to disrupt the fred - like shape ,
@xmath183 s must vary by a factor of at least 2 from what they would be if everything was constant with @xmath23 .
consider the effect of different times of emission ( i.e. , different @xmath177 s but similar @xmath180 s ) . the @xmath177 s would have to vary by about a factor of 2 to change @xmath183 s by a factor of 2 . consider constant @xmath177 s but different @xmath180 s . to spread out out the emitting regions so their distance from us varies by about @xmath8 requires that @xmath184 or that the radius of the shell is @xmath185 the @xmath186 and @xmath187 would have to vary by about @xmath188 .
such differences could arise from different baryon loading as a function of @xmath23 .
there are two observations that argue against this explanation .
first , even the bursts with fred - like envelopes show low filling factors ( see the solid squares in fig .
second , if the @xmath186 s differ by @xmath188 , then peak near the end of the burst should have widths that are at least a factor of 2 wider .
bursts normally do not appear to have such a trend .
this explanation requires an energy reservoir like a uniform shell : @xmath165 . in conclusion , both the gamma - ray phase and the x - ray afterglow show similar rapid variability , both have values of @xmath189 that are similar .
such variability implies that either grbs are central engines , or that the local spherical symmetry is broken on a scale much smaller than @xmath2 . in the last case
, the structure can either be due to variations in @xmath3 as a function of @xmath23 ( perhaps due to small scale variations in the baryon loading , explanation 6 above ) , or due to only a fraction @xmath5 of the shell becoming active ( explanations 3 , 4 , 5 above ) .
the required total energy is @xmath190 , @xmath191 and @xmath192 for explanations 3 , 4 , and 5 , respectively .
using typical values , the total energy can vary from @xmath174 to @xmath193 erg .
we note that explanation 5 ( a jet much narrower than @xmath2 ) is the only explanation that can also easily explain gaps in the time history .
we thank cheng ho and chuck dermer for useful comments .
jay norris provided the time histories analyzed for figure 3 .
this work was done under the auspices of the us department of energy and was funded in part by the nasa gro guest investigator program .
rccccc 143 & 1.5 & 1.10 & 6.7 & 2.43 & 0.812 219 & 3.0 & 0.51 & 6.3 & 2.37 & 1.184 249 & 1.8 & 1.43 & 6.7 & 3.75 & 1.312 257 & 42.7 & 0.50 & 28.5 & 7.21 & 2.181 394 & 2.8 & 0.76 & 37.0 & 0.55 & 0.157 467 & 6.7 & 0.50 & 8.8 & 3.65 & 0.956 469 & 5.2 & 0.26 & 4.0 & 3.20 & 0.450 563 & 79.8 & 0.50 & 8.6 & 44.41 & 7.823 647 & 4.7 & 0.90 & 16.8 & 2.43 & 1.636 660 & 2.7 & 0.26 & 5.5 & 1.22 & 0.633 676 & 2.8 & 0.76 & 18.4 & 1.11 & 0.668 678 & 4.1 & 0.32 & 20.5 & 0.62 & 0.185 829 & 3.4 & 0.38 & 6.0 & 2.10 & 1.740 841 & 2.1 & 0.45 & 5.8 & 1.54 & 0.708 1122 & 2.9 & 0.51 & 6.1 & 2.38 & 0.979 1141 & 10.0 & 1.54 & 6.9 & 21.47 & 8.346 1288 & 0.6 & 0.89 & 100.2 & 0.05 & 0.016 1440 & 0.4 & 0.51 & 2.8 & 0.69 & 0.230 1541 & 1.0 & 0.48 & 10.6 & 0.44 & 0.255 1606 & 2.9 & 0.51 & 53.4 & 0.27 & 0.088 1625 & 3.7 & 0.47 & 10.0 & 1.67 & 0.557 1652 & 2.6 & 0.83 & 23.9 & 0.86 & 0.468 1663 & 3.9 & 1.02 & 10.6 & 3.61 & 1.370 1676 & 1.3 & 0.32 & 30.8 & 0.13 & 0.032 1815 & 2.5 & 0.32 & 10.8 & 0.72 & 0.249 1883 & 120.2 & 0.50 & 3.6 & 100 & 13.420 2067 & 2.4 & 0.64 & 6.9 & 2.14 & 0.918 2080 & 4.0 & 0.32 & 22.7 & 0.55 & 0.328 2083 & 2.4 & 0.38 & 7.6 & 1.18 & 0.319 2090 & 0.3 & 0.26 & 8.6 & 0.10 & 0.039 2110 & 3.6 & 0.70 & 7.5 & 3.22 & 0.475 2138 & 0.9 & 0.77 & 12.0 & 0.57 & 0.214 2156 & 1.1 & 0.78 & 92.0 & 0.09 & 0.031 2213 & 0.3 & 0.90 & 51.6 & 0.05 & 0.013 2228 & 1.5 & 0.38 & 30.3 & 0.19 & 0.038 2329 & 2.5 & 0.70 & 7.4 & 2.32 & 0.786 2436 & 0.9 & 0.90 & 24.6 & 0.32 & 0.197 2450 & 2.6 & 0.26 & 11.9 & 0.54 & 0.377 2533 & 1.8 & 0.45 & 26.0 & 0.30 & 0.127 2586 & 1.3 & 0.26 & 7.9 & 0.40 & 0.143 2700 & 1.5 & 0.32 & 32.6 & 0.14 & 0.035 2798 & 3.8 & 1.02 & 9.0 & 4.19 & 0.599 2812 & 0.5 & 0.38 & 14.1 & 0.14 & 0.036 2831 & 1.5 & 0.51 & 66.0 & 0.11 & 0.037 2855 & 1.8 & 0.83 & 13.2 & 1.11 & 0.344 2889 & 1.3 & 0.38 & 20.6 & 0.24 & 0.047 2929 & 2.4 & 0.19 & 14.1 & 0.31 & 0.178 2984 & 1.5 & 0.26 & 11.8 & 0.31 & 0.117 2993 & 8.4 & 0.26 & 17.7 & 1.17 & 0.574 3035 & 4.0 & 0.58 & 35.8 & 0.62 & 0.109 3057 & 3.7 & 0.77 & 13.4 & 2.05 & 1.569 3128 & 2.2 & 0.26 & 14.1 & 0.39 & 0.106 970828 & & 7200 & @xmath194 & 0.10 & = --2 costa , et al .
, 1997 , nature , 387 , 783 dar , a. , 1998 , submitted to apjl , astro - ph/9709231 fenimore , e. e. , madras , c. d. , & nayakshin , s. , 1996 , apj 473 , 998 , astro - ph/9607163 fenimore , e. e. , ramirez - ruiz , e. , & sumner , m. c. , 1998 , gamma - ray bursts : 4th huntsville symposium , eds .
meegan , preece , & koshut , ( aip proc .
428 ) pg 657 , astro - ph/9712303 fenimore , e. e. , & sumner , m. c. , 1997 , all - sky x - ray observations in the next decade , eds .
matsuoka , m. , & kawai , n. , pg 167 , astro - ph/9705052 frail , d. , kulkarni , s. r. , nicastro , l. , feroci , m. , & taylor , g. b. , 1997 , nature , 389 , 261 kobayshi , s. , piran , t. , & sari , r. , 1997 , apj , 490 , 92 meegan , c. a. , et al .
1996 , apjs , 106 , 65 , p. , & rees , m. j. , 1993 , apj , 405 , 278 , p. , & rees , m. j. , 1997 , apj , 476 , 232 , p. , rees , m. j. , & wijers , r. a. m. j. , 1998 , apj submitted , astro - ph/9709273 metzger , m. r. , et al .
1997 , nature , 387 , 878 norris , j. p. , et al .
, 1996 , apj 459 , 393 panaitescu , a. , & , p. , 1998
, apj submitted piran , t. , & sari , r. , 1997 , apj 485 , 270 piro , l. , et al . , 1997 , a&a submitted rees , m. j. , & , p. , 1994
, apj , 430 , l93 rees , m. j. , & , p. , 1998 ,
apjl , submitted , astro - ph/9712252 reichart , d. , 1997 , apj , 485 , l57 sari , r. , & piran , t. , 1997a , apj in press sari , r. , & piran , t. , 1997b , mnras in press sari , r. , piran , t. , & narayan , r. , 1998 , preprint astro - ph/9712005 sumner , m. c. , & fenimore , e. e. , 1998 , gamma - ray bursts : 4th huntsville symposium , eds .
meegan , preece , & koshut , ( aip proc .
765 , astro - ph/9712302 tavani , m. , 1997 , apj , 483 , l87 waxman , e. , kulkarni , s. r. , & frail , d. a. , 1998 , apj , submitted , astro - ph/9709199 wijers , r. a. m. j. , rees , m. j. , & , p. , 1997 ,
mnras , 288 , l51 yoshida , a. , et al . , 1998 , gamma - ray bursts : 4th huntsville symposium , eds .
meegan , preece , & koshut , ( aip proc .
428 ) pg 441 | the variability observed in many complex gamma - ray bursts ( grbs ) is inconsistent with causally connected variations in a single , symmetric , relativistic shell interacting with the ambient material ( `` external shocks '' ) .
rather , either the central site must produce @xmath0 erg s@xmath1 for hundreds of seconds ( `` internal shocks '' ) or the local spherical symmetry of the shell must be broken on an angular scale much smaller than @xmath2 where @xmath3 is the bulk lorentz factor for the shell . the observed variability in the external shock models arises from the number of causally connected regions that ( randomly ) become active .
we define the `` surface filling factor '' to be the ratio of the area of causally connected regions that become active to the observable area of the shell . from the observed variability in 52 batse bursts
, we estimate the surface filling factor to be typically @xmath4 although some values are near unity .
we find that the surface filling factor , @xmath5 , is @xmath6 in both the constant @xmath3 phase ( which probably produces the grb ) and the deaccelerating phase ( which probably produces the x - ray afterglows ) . here
, @xmath7 is a typical time scale of variability and @xmath8 is the time since the initial signal .
we analyze the 2 hr flare seen by asca 36 hr after the grb and conclude that the surface filling factor must be small ( @xmath9 ) in the x - ray afterglow phase as well .
compared to the energy required for an isotropic shell , @xmath10 , explanations for low surface filling factor can either require more energy ( @xmath11 erg ) or less energy ( @xmath12 erg ) .
thus , the low filling factor can not be used as a strong argument that grbs must be internal shocks .
= 1 # 1#2#1 |
there is now a lot of interest in the connection between high energy physics and cosmology .
it is motivated by the possibility that processes not reproducible here on the earth actually occurred in the early universe .
for this reason , a lot of work is currently in progress to predict in detail the traces that such processes could have left , in order to recognize them and gain insight into physics that is still unknown , or only theoretically approached .
the unknown sector of physics extends from the energy scales presently explored by accelerators , described successfully by the standard model of the known elementary particles , up to the scales of more fundamental theories , perhaps supersymmetry and supergravity ; such regimes , as thought in the early universe , should have taken place at temperatures @xmath0 ( in energy units ) in the interval @xmath1 according to our hypotheses , two main classes of phenomena took place in the early universe : an era of accelerated expansion , the inflation , and the breaking of high energy symmetries , see @xcite .
the first process should leave traces in the form of gaussian and scale - invariant density fluctuations @xcite ; this visually corresponds to a completely disordered distribution of hills and wells in the density field , covering all the scales .
the second process leaves completely different traces : spatially limited structures , like topological defects @xcite or bubbles made both of true and false vacuum @xcite . at the present status of the theoretical knowledge
, their production may occur with or without inflation .
models able to produce such structures both during and after inflation have been studied @xcite . in order to be observable ,
the first case is most interesting , since the size of the structure is stretched by the stage of inflation after their formation , up to cosmologically interesting scales of tens of comoving mpc or more . as well as the gaussian fluctuations ,
these structures may be considered as seeds for the cmb perturbations . in the recent past , they have been thought as candidates for the structure formation process with preliminary discouraging results @xcite , even if the numerical simulations and the models to explore are far to be exhausted ; unfortunately , we do not have a good theory to predict their exact properties and abundance .
the only sure thing is that the detection of at least one of them would be the first observational evidence of the existence of high energy symmetries .
so the analysis here regards the signal from each single seed , without requiring neither that they dominate the structure formation process , nor that their signature is present on the whole sky cmb power spectrum .
these seeds may also be thought to possess some spatial symmetries , both because appropriate and because the problem becomes simpler .
spherical and cylindrical symmetries are particularly simple and appropriate for bubbles , monopoles and strings , also forming loops @xcite ; also they allow to write simple and suitable formulas for the cmb perturbations ; we refer to @xcite for a more quantitative and detailed exposition of these aspects .
in this work we point out the characteristic signature of these structures on the cmb , in direct connection with the forthcoming whole sky cmb experiments @xcite .
as we shall see , their spatial shape combined with the undulatory properties of the cmb physics mix and produce their unambiguous signs .
we begin with some necessary technical detail , but we hope to finish with physically simple and intuitive results . in linear theory , and assuming a friedmann robertson walker ( frw ) background , the equations driving the energy density perturbation and the peculiar motion of photons can be obtained from the linearized einstein equations @xcite .
perturbations may be classified as scalar , vector and tensor with respect to spatial rotations ; bubbles or topological defects are essentially fluctuations in the energy density composed by matter , radiation as well as scalar fields , therefore the case of interest here is the first one .
the linearization implies a gauge freedom with respect to infinitesimal frame transformations ; we choose the newtonian gauge which physically corresponds to observers at rest with respect to the comoving expansion and experiencing the latter isotropic @xcite .
perturbations in the cmb photons are coupled to the fluctuations of the other constituents of the cosmic energy density . in particular
, thomson scattering between baryons and photons induces polarization perturbations in the cmb , being an anisotropic process @xcite . at early times , the fluid is so dense that the photons free path @xmath2 vanishes ; it is small with respect to the hubble horizon @xmath3 and the perturbation wavelength @xmath4 @xcite .
therefore , the cmb equations may be expanded in powers of @xmath5 and @xmath6 . in practice ,
the first order terms become important at decoupling , when the photons free path suddenly increases .
one can consider cmb photons traveling on a direction @xmath7 in the spacetime point @xmath8 , where @xmath9 is the conformal time defined in terms of the ordinary time @xmath10 and of the scale factor @xmath11 by @xmath12 .
cmb temperature and polarization perturbations are expanded into spherical harmonics describing the dependence on @xmath7 .
this treatment was firstly used in @xcite and recently expanded to include non - flat geometries and non - scalar perturbations @xcite . for each fourier mode ,
computations are performed in the @xmath13frame , where the wavevector @xmath14 is the polar axis for the angular expansion ; the fixed laboratory frame is instead indicated as the @xmath15frame ; this distinction is particularly important for the perturbations considered here @xcite . to fix the ideas , before decoupling the cmb dynamics
may be considered at the zeroth order in the thomson scattering terms .
thus the equations for the energy density fluctuations @xmath16 and peculiar motion @xmath17 of photons are easily gained by the linearized conservation equations and have the simple following form : @xmath18 @xmath19 @xmath20 and @xmath21 are the two scalar metric perturbations accounting for fluctuations from all the fluid species @xcite , and @xmath22 are the cosmic scale factor and its derivative with respect to the conformal time .
the terms containing the unperturbed baryon density @xmath23 are a residual of the coupling between photons and baryons , at the zeroth order in the thomson scattering .
also we point out that @xmath24 are simply linked to the monopole and the dipole of the cmb temperature fluctuation dependence on the photon propagation direction @xmath7 @xcite : @xmath25 equations ( [ rho - v ] ) may be put together in the following most simple form : @xmath26 @xmath27 this is a wave equation with friction of cosmological origin ( proportional to @xmath28 ) and forcing , gravitational terms at the second member .
focus on the friction term .
its effect is simple .
much before the horizon crossing the following conditions are satisfied @xcite : the second term of the first member dominates over the third one , the gravitational potentials are constant and @xmath29 ; therefore the solution is trivially @xmath30 and no friction effect exists at all . at the horizon
crossing the last term of the first member becomes important , and @xmath16 starts to oscillate , making @xmath31 not vanishing . thus the friction becomes active and damps the oscillations .
now we come to the central arguments of the present work . differently from analogous problems in cosmology ,
the friction term is multiplied by @xmath32 . in most cosmological models
the baryon content is of the order of percent with respect to the dark matter component , because of the severe constraints from nucleosynthesis @xcite .
therefore it is evident that @xmath33 between equivalence and decoupling , reducing substantially the friction term for the oscillations occurring at these epochs . for the perturbations considered here , the forcing terms in ( [ rhov ] ) are active on spatially limited regions , occupied by the seed .
the arguments above show that outside the seed the equation for @xmath16 is @xmath34 this says that the oscillations occurring at the horizon crossing are brought _ outside _ the seed with the cmb sound velocity .
they shall reach the sound horizon at the time considered , given by @xmath35 and it can be shown that it corresponds roughly to one degree in the sky ; we remark that @xmath16 waves mean @xmath36 waves from ( [ sb ] ) , and this means polarization waves , since polarization and temperature perturbations are tightly coupled @xcite .
the consequences of the exposed arguments are straightforward .
consider a spatially limited seed intersecting the last scattering surface .
its cmb signal is made of waves extending approximatively over one degree in the sky .
this is extremely interesting for the future whole sky , high resolution observations @xcite .
these same waves are a unambiguous proof of the fact that really the seed was generated well before decoupling , simply because any other astrophysical source does not produce them because it formed after decoupling , appearing point - like .
these concepts might be observationally decisive if our theoretical thinking is quite right and high energy symmetry breaking traces are really present in the nearby universe . in the next section
we give a numerical example of the realization of the phenomenology exposed here in the context of the standard cosmological scenario .
we refer to @xcite for all the computational and formal details .
we considered spherical and infinitely long cylindrical energy concentrations in a background formed by cold dark matter @xmath37 , baryons @xmath38 , photons and massless neutrinos , assumed distributed initially adiabatically . at any time , and for this kind of sources , the expressions of the cmb temperature and polarization perturbations may be written in simple and intuitive forms @xcite .
we give here the expressions for spherical sources . *
temperature*. the temperature cmb perturbation at any spacetime point @xmath39 for a spherical source is given by @xcite : @xmath40 @xmath41 and @xmath42 are the legendre polynomials and the fractional order bessel function respectively .
note how the spatial directions @xmath7 and @xmath43 , describing geometrically the problem , are _ outside _ the fourier integral . in spite of the virtually infinite series ,
already the @xmath44 term is the substantial component of the signal , as we show below . * polarization*. it is described by the stokes parameters @xmath45 and @xmath46 on the plane perpendicular to the photon propagation direction @xmath7 . as for the temperature case , each fourier mode of the polarization perturbations admit an expansion in spherical harmonics with coefficients @xmath47 and @xmath48 . remembering that these quantities are defined in the @xmath13frame , the problem may be further simplified since @xmath49 for scalar perturbations
. also we mention that as a distinction with respect to the temperature case , the _ tensor _ spherical harmonics are now required , accounting for the tensor properties of the polarization ; for this reason , @xmath47 and @xmath48 are defined for @xmath50 @xcite .
the polarization cmb perturbation at any spacetime point @xmath39 in presence of a spherical source is given by @xcite : @xmath51 @xmath52 @xmath45 and @xmath46 are essentially identical except for the geometrical dependence on @xmath53 , that is the angular coordinate of the projection of @xmath54 on the plane perpendicular to @xmath7 , as seen in the @xmath15frame : if the latter is chosen so that @xmath55 we have @xmath56 .
note that now the _ second order _
legendre polynomials compare into the sum .
they are meaningful since prevent photons traveling radially ( @xmath57 ) to be polarized ; this is correct since in that case no preferred axis exists for polarization . as for the temperature case , the substantial component of the perturbation is given by the first term of the sum , @xmath58 .
note that these relations are _ independent _ on the nature of the seed , which is encoded in the @xmath59 coefficients and could be made of matter and radiation as well as scalar field , that has the only restriction in its sphericity .
each coefficient @xmath59 obeys motion equations , and together with the linearized einstein ones the whole system may be evolved in time , see @xcite and references therein
. figures [ lf1 ] and [ lf2 ] show the time evolution of the cmb temperature and polarization perturbations . in each panel
the perturbation profile is shown as a function of the distance from the seed center and symmetry axis respectively for the spherical and cylindrical case .
the perturbations are normalized with the density contrast at decoupling @xmath16 , taken at the center for the spherical seed and on the symmetry axis for the cylindrical one . at the horizon
crossing , cmb waves form and travel outward , just like the waves from a pebble thrown in a pond .
the upper panels regards the temperature signals , while the lower ones shows the polarization amplitude .
the thin lines represents the signal only from the monopole term in ( [ dtt ] ) , while the thick ones contains all the contributions ; this shows how the monopole is the dominant term .
as indicated , photons travel perpendicularly to the radial distance @xmath43 for the spherical case and on the plane perpendicular to the symmetry axis for the cylindrical case .
the symmetries of the seed allow to choose the @xmath15frame axes so that the polarization is given by a pure @xmath45 term @xcite .
in the spherical case they are simply parallel and perpendicular to the plane formed by @xmath7 and the radial direction @xmath43 in the scattering point ; in the cylindrical case they are the symmetry axis @xmath60 and the direction perpendicular to @xmath7 and @xmath60 .
the waves at any time occupy the position of the cmb sound horizon ( [ hs ] ) , and have roughly the same amplitude of the perturbation inside the seed .
really , in the polarization case , they are the very dominant component of the signal .
also a marked correlation is evident between the polarization and temperature waves .
the symmetries of the seeds constrain the signal .
photons traveling radially in a spherical seed must be not polarized since this problem is symmetric with respect to rotations around the propagation direction @xmath7 . in the cylindrical case
the symmetry axis itself is a preferred direction and polarization perturbations may affect photons traveling away from the symmetry axis .
these figures represent the cmb perturbations around spatially limited seeds at different times .
now we show in the spherical case the real simulation of the cmb anisotropies , in order to show how the exposed behavior is maintained @xcite .
figure [ lf3 ] shows the cmb anisotropy from the spherical seed if it intersects the last scattering surface , as a function of the angle @xmath62 from the center .
the sachs - wolfe effect regarding the zone physically occupied by the seed , @xmath63 has been included .
the cmb perturbation waves have been photographed by the decoupling photons . of course
the signal is a function of the relative disposition of the seed with respect to the last scattering surface .
different cases have been displayed : the solid line represents the case in which the seed center lies exactly on the last scattering surface ; the dashed and dotted dashed lines shows the signals if the seed lies 30@xmath64 mpc within or outside our hubble sphere . in particular ,
if the distance between seed and last scattering surface is much larger then a cmb sound horizon at decoupling , roughly 100@xmath64 mpc , the seed is visible only through the integrated sachs - wolfe effect if it lies within our hubble sphere ; in this case however the cmb waves could not be detected , even if distinctive signals would arise @xcite .
the amplitude of the signal is roughly as expected for linear perturbations with density contrast @xmath16 and size @xmath65 smaller than @xmath3 at decoupling : @xmath66 and a few percent of this for the polarization amplitude .
of course a lot of work has to be done to predict in detail the signs from any relic of the early universe ; at the moment such prediction exists only for the temperature signals from inflationary bubbles @xcite . the final general statement that we make here
is the following .
if the forthcoming high resolution cmb maps should contain perturbed spots surrounded by anisotropy waves reaching the distance of the cmb sound horizon at decoupling , then we could conclude that the processes that generated such signals belong to the unknown very high energy sector of physics .
this would open new possibilities for testing fundamental physical theories .
lyth d.h . &
riotto a. 1998 , to appear in physics reports , hep - ph/9807278 mukhanov v.f .
, feldman h.a . &
branderberger r.h .
1992 phys.rep .
215 , 203 bardeen j.m .
1980 , phys.rev.d22 , 1882 ; kodama i. & sasaki m. 1984 , progr .
of theor.phys.supp .
78 , 1 ma c.p . & bertschinger e. 1995 , ap.j .
455 , 7 kosowsky a. 1996 , annals phys .
246 , 49 , astro - ph/9501045 hu w. , seljak u. , white m. & zaldarriaga m. 1998 , phys.rev.d57 , 3290 , astro - ph/9709066 ; hu w. & white m. 1997 phys.rev.d 56 , 596 , astro - ph/9702170 baccigalupi c. 1998 , submitted to phys.rev.d , astro - ph/9811176 baccigalupi c. 1998 , ap.j 496 615 , astro - ph/9711095 ; amendola l. , baccigalupi c. & occhionero f. 1998 ap.j . lett .
492 , l5 bouchet f. , bennet d. & stebbins a. 1988 , nat .
355 , 410 coleman s. & de luccia f. 1980 , phys.rev.d 21 , 3305 ; lee k & weinberg e.j .
1987 , phys.rev.d36 , 1088 olive k.a .
1999 , nucl.phys.proc.suppl .
70 521 , astro - ph/9712160 branderberger r.h .
1998 , invited lectures at whepp-5 , iucaa , pune , india , pramana 51 , 191 , hep - ph/9806473 vilenkin a. 1985 , phys.rep .
121 , 263 basu r. , guth a.h . & vilenkin a. 1991 phys.rev.d44 , 340 yokoyama j. 1988 , phys.lett.b 212 , 273 kofman l. , linde a. & starobinsky a. 1997 , phys.rev.d 56 , 3258 planck surveyor home page : + http://astro.estec.esa.nl/sa-general/projects/planck/ ; + micowave anisotropy probe ( map ) home page : http://map.gsfc.nasa.gov | we consider anisotropies in the cosmic microwave background ( cmb ) generated by spatially limited seeds ; these objects could correspond to relics of high energy symmetry breaking in the early universe .
it is shown how the cmb perturbation propagate beyond the size of the seed in the form of waves traveling with the cmb sound velocity . moreover , these waves are the substantial part of the signal , both for polarization and temperature .
the explanation of this phenomenology in terms of the cmb equations is given .
observationally , this effect is threefold promising .
first , it enlarges the signal from a seed intersecting the last scattering surface to the scale of the cmb sound horizon at decoupling , that is roughly one degree in the sky .
second , it offers cross correlation possibilities between the polarization and temperature signals .
third , it allows to unambiguously distinguish these structures from point - like astrophysical sources . |
planetary evaporation is one of the most crucial factors that determines the evolution and fate of close - in planets .
the atmospheric escapes of hot jupiters have been extensively investigated from both the extreme ultraviolet observations and theory .
the impacts of evaporation on the evolution of even smaller planets , including super - earths and super - neptunes , has been also recognized .
recently @xcite ( hereafter r12 ) found violent variation of transit depths and an ingress - egress asymmetry of kic 12557548 , which have been interpreted as evidences of a catastrophic evaporation of a small planet ( r12 , * ? ? ?
* hereafter pc13 ) . analyzing the intensity of forward scattering , and
@xcite found that the dust expected as an occulter of kic 12557548is consistent with sub - micron grains .
r12 and pc13 estimated the mass loss rate of kic 12557548b , as @xmath0 .
however , the dominant energy source of the enormous evaporation remains unknown .
one proposed possible scenario is that high bolometric energy input at the semi - major axis @xmath5 au itself produces the hydrodynamic escape .
pc13 have constructed a radiative - hydrodynamic model of kic 12557548bassuming a parker - type wind driven by hot planetary surface with the equilibrium temperature @xmath1 2000 k. they concluded that a planet with mass @xmath6 can account for the mass loss rate of kic 12557548b .
the hydrodynamic escape driven by x - ray and ultraviolet ( xuv ) radiations as has been discussed for hot jupiters is another candidate for the energy source .
however , an xuv observation of the transit is impossible for kic 12557548with current satellites due to its distance . instead , we focus on the time - series of the transit depth . in this letter , we reanalyze the _ kepler _ data of kic 12557548 , but for longer period @xmath1 3.5 yr , than that analyzed by r12 .
we find the correlation between the transit depth and the modulation of the light curve using starspots .
these results indicate that the evaporation of kic 12557548bresults from some energy related to the stellar activity , rather than the bolometric one .
we analyze archived long cadence data of kic 12557548between quarter 1 and 15(q1-q15 ; @xmath1 3.5 yr ) using the @xmath7 package to process the data @xcite .
figure [ fig : lc ] shows the cotrended long time cadence data of kic 12557548(panel a ) and the flattened light curve ( panel b ) , extracted by _ kepcotrend _ and
_ kepflatten_. the time series of the transit depth is obtained by averaging three bins ( = 1.47 hr ) around the central time of transit .
we adopt the central time as @xmath8 , where @xmath9 days ( r12 ) and the @xmath10 days .
our definition of the transit depth is the average of the 3 bins of the flatten light curve , as shown by red points in figure [ fig : lc](b ) . as shown in figure [ fig : lc](c ) , the distribution of the flatten light curve after removal of the transit periods is well fitted by a normal distribution , at least for the lower side , which verifies that our extraction method could extract almost all of the transit periods .
figure [ fig : tradep ] shows the time series of the extracted transit depth ( top ) with the cotrended light curve of the star ( bottom ) . ]
figure [ fig : period ] shows lomb - scargle periodograms of the time series of the transit depth .
we find three peaks above false alarm probability ( fap ) = 0.1 % , @xmath3 days,@xmath11 days , and @xmath12 days , corresponding to fap= 0.04 % , @xmath13 % , and @xmath14 % .
we also compute periodogram of the cotrended light curve excluding the detected transit duration ( bottom panel in figure [ fig : period ] ) .
we find the most prominent peak at @xmath4 days and its harmonic peaks .
we also performed high resolution spectroscopy for kic 12557548 with the high dispersion spectrograph on the subaru telescope .
we adopted the standard `` ra '' setup with the image slicer no.2 ( @xmath15 and @xmath16 nm ) .
the basic spectroscopic parameters were extracted based on @xcite and @xcite : @xmath17 k , @xmath18 , @xmath19=0.09\pm0.09 $ ] , and @xmath20 km s@xmath21 .
the resultant systematic error in @xmath22 is quite uncertain because of the faintness of the target star ( @xmath23 ) , but we can safely rule out a large rotational velocity , @xmath24 km s@xmath21 , corresponding to a rapid rotation @xmath25 days ( we assume @xmath26 ; r12 ) . on the assumption that this value is the stellar rotation period , the shortest periodicity found in the transit depth variation @xmath3 days is consistent with the stellar rotation period .
we concentrate on the periodicity of @xmath27 through the rest of the paper .
we fold the transit depth variation with @xmath27 ( the top and middle panels in figure [ fig : lcfold ] ) .
the mean value of binned data has @xmath1 30 % variation .
we also fold the cotrended light curve with @xmath27 ( the bottom panel ) and find that the folded cotrended light curve is negatively correlated with the folded depth variation . a large starspot
can survive for many years @xcite .
the 2 % variation of the folded light curve can be interpreted as long term variation due to a large starspot associated with a local active area .
hence our interpretation of the anti - correlation is that the planet tends to make deeper occultation when facing the large starspot .
in general , stellar visibility depends on competition among starspots and faculae . in the case of the sun
, spot modeling predicts that a single active area decreases the visibility only for the angle between the line - of - sight and the local active area on the stellar surface @xmath28 due to the large faculae - to - spots ratio @xmath29 .
however , for several stars with large flux modulation , smaller values of @xmath30 were estimated from the spot modeling and faculae for these stars do not contribute significantly to the visibility . throughout the rest of the paper , we assume that the local active area decreases stellar visibility . the stellar visibility variation results in the apparent variation of the measured depth via normalization . considered this effect for different epochs of stellar activity ( a @xmath1 1 year interval ) .
we use a similar formalism although we consider this effect during one stellar rotation .
the apparent variation of the depth @xmath31 is related to that of the visibility during one revolution , @xmath32 as @xmath33 where @xmath34 and @xmath35 are the measured depths when the active area is located on the side visible to us and the opposite side , @xmath36 and @xmath37 are the visibilities for the two side . denoting the average surface brightness of photosphere by @xmath38 and @xmath39 for the two sides and
that along the trajectory of transit by @xmath40 , @xmath41 , we can express @xmath42 where @xmath43 and @xmath44 .
assuming that the spot crossing is negligible ( @xmath45 ) and @xmath46 , we obtain @xmath47 .
if the surface brightness outside the spot areas is homogeneous , we obtain @xmath48 . in our case , we expect the 0 - 4 % variation of @xmath49 to be anti - correlated to the stellar visibility .
since the observed variation is @xmath50 30 % , we conclude that the periodicity of the depth variation synchronized with the stellar rotation is not a false positive .
@xcite found seasonal variation of transit depth for hat - p-7 b due to instrumental systematics .
they did not report any other periodicity @xmath51 1 yr . since the periodicities @xmath52 and @xmath53 are larger than period of a quarter , instrumental systematics should be considered in further studies .
to consider possible scenarios of evaporation , two important constraints from observation , the mass - loss rate @xmath54 and the planet radius @xmath55 , should be taken into account .
the mass - loss rate was constrained by both r12 and pc13 in similar but slightly different ways , @xmath56 and @xmath57 . here
we summarize the essential part of their derivation .
the @xmath54 can be estimated using the total cross section of transit @xmath58 ( we adopt @xmath59 ) , dust density @xmath60 , typical length of optical path @xmath61 and typical time scale @xmath62 , @xmath63 where @xmath64 is the gas - dust ratio . r12
adopted @xmath65 hr , assuming the typical velocity of outflow @xmath66 km / s , @xmath67 , and @xmath68 as a typical grain size , which yields @xmath69 .
pc13 uses an orbital period of @xmath70 hr , @xmath71 , @xmath72 , which results in @xmath73 .
the grain size estimated from the forward scattering supported @xmath74 .
we computed the auto correlation of the @xmath75 and found a flat correlation for a time interval , @xmath76 days .
it supports @xmath77 . assuming that the dust density is an order of @xmath78 , we use @xmath79 as a fiducial constraint of the mass loss rate .
another constraint is the upper limit of @xmath55 .
estimated the 1 @xmath80 upper limit of the maximum planetary radius as @xmath81 translating photometric accuracy of the phase folded light curve of no - transit - like feature ( around the period in red in figure [ fig : tradep ] ) into the radius . interpreting the starspots - depth relation in a straight forward fashion
, the atmospheric escape rate should correlate with the area of the active area on the stellar surface . because the optical flux does not change much ( @xmath1 2% ) and is negatively correlated with the transit depth , the bolometric radiation is unlikely to be a source of the relation .
strong xuv radiation is often related to stellar spots since the xuv flares originate from magnetic loops .
the x and/or uv periodicity with rotation were found in several stars . while the sunspots exhibit much smaller variation than kic 12557548 , an order - of - magnitude of variability in solar x - ray has been observed over a month . we found a sudden 6 % increase in the flux during 30 minutes similar to the flare in the short cadence data .
such a strong flare will create strong xuv variation in a timescale of hours , which might account for the transit depth variation in every transit .
however , r12 questioned whether the input energy is sufficient to account for the mass loss rate .
we revisit this problem using the theoretical mass - radius relation of sub / super - earths with several different compositions . because of the lack of xuv observation , we manage to estimate the xuv flux of kic 12557548from the x - ray flux - rotation relation , which provides @xmath82 erg / s for @xmath83 days .
considering the fact that kic 12557548has a very large active area that creates 2 % continuous flux variation over 3.5 yr , we adopt @xmath84 , corresponding to @xmath85 erg / s . in the energy - limit regime
, the mass loss rate can be estimated as @xmath86 where @xmath87 is the radius where gas is bound to the planet , @xmath88 is the radius where the bulk of incoming xuv flux is absorbed , @xmath89 is the efficiency for converting the input energy to the kinetic energy of the atmosphere . because the base level of the xuv deposition @xmath88 is under discussion
, we adopt the radius where optical depth @xmath90 conservatively , i.e. @xmath91 . a correction factor for the roche lobe effect , @xmath92 , is given by , @xmath93 where @xmath94 is the hill radius , @xmath95 on the basis of @xcite , we compute the mass - radius relation of planets composed of 100 % water , 50% water+50% rocky core , 10% water+90 % rocky core ( volatile - rich planets ) , 100 % rock , and 100 % iron assuming the equilibrium temperature of kic 12557548b(@xmath1 1800 k ) .
we neglect the contribution of a thin atmosphere , which should be evaporating , for the 100 % rock and iron planets to compute the relation .
the interior structure of the planet is computed with equations of hydrostatic equilibrium , mass conservation , and energy conservation ( e.g. * ? ? ? * ) and the equation of state .
a detailed description is found in @xcite .
figure [ fig : mlr ] ( left ) shows the mass - radius relation of these planets .
we find that the radius of volatile - rich ( the water fraction being @xmath96 10 % ) planets can not be smaller than the upper limit of planetary radius @xmath97 because the gravity can not bind the atmosphere
. therefore kic 12557548bis unlikely to be a volatile - rich planet . figure [ fig : mlr ] ( right ) shows the @xmath54 expected in the energy - limit case , assuming the relatively high flux @xmath98 and an assumption for the efficiency of @xmath99 .
the density of the planet composed of iron is too dense to obtain the mass loss rate needed .
the @xmath54 of the rocky planet is @xmath100/gyr .
the roche lobe effect increases @xmath54 to two times larger than that without the tide .
r12 discussed that the xuv luminosity is at least a factor of 10 too low compared to the luminosity needed assuming @xmath101 .
the roche lobe effect and the assumption of @xmath89 are likely to cause the difference by one order - of - magnitude in @xmath54 between r12 and us .
hence , a rocky planet is possible as a candidate for kic 12557548b .
one might consider that volatile components on the rocky planet increase the planetary effective radius and @xmath54 .
we derive the surface pressure needed to increase the transit radius @xmath102 to @xmath103 times of the radius of the silicate core @xmath104 ( @xmath105 ) though the observed transit radius is not identical to the effective radius for equation ( [ eq : energylimited ] ) .
we adopt @xmath106 , i.e. @xmath107 times larger cross section . the atmospheric height from the ground to the point where the optical depth through the planetary limb is unity is estimated by assuming an isothermal atmosphere with temperature @xmath108 and a power - law dependence of the opacity on pressure ( @xmath109 ; * ? ? ?
* ) , @xmath110 where @xmath111 is the pressure scale height , @xmath112 , @xmath113 and @xmath114 are the surface pressure at the lower boundary , mean - molecular weight and proton mass .
the vertical optical depth @xmath115 is linked to the optical depth of the chord ( e.g. * ? ? ? * ; * ? ? ?
* ; * ? ? ?
* ) as @xmath116 it yields @xmath117 where @xmath118 is the visible opacity at the surface . then we obtain @xmath119 } \end{aligned}\ ] ] where @xmath120 is the escape parameter and @xmath121 for @xmath122 and @xmath123 . for the na atmosphere , which is a possible dominant component of the atmosphere of the rocky planets ( e.g. * ? ? ?
* ) , @xmath124 for @xmath125 bar , where we computed @xmath126 based on atomic line data @xcite with the method by @xcite and took the logarithm average of @xmath126 ( see eq . [ [ eq : hhh ] ] ) for the _ kepler _ band . adopting @xmath127
, we obtain @xmath128 bar and @xmath129 bar for @xmath130 ( @xmath131 ) and @xmath132 ( @xmath133 ) , which are much higher than the equilibrium vapor pressure of the olivine , @xmath134 bar for 2000 - 2500 k @xcite ( see also * ? ? ?
thus the na atmosphere vaporized at the equilibrium temperature is unlikely to increase the cross section of @xmath135 .
though we assumed the equilibrium temperature as the surface temperature for simplicity , further study of the atmospheric structure over magma ocean is needed to derive the realistic contribution of the volatile components .
star - planet magnetic interaction is another possible scenario for a catastrophic evaporation that correlates with stellar activity .
recently @xcite proposed a new scenario in which energetic electrons released by magnetic reconnection at the boundary of the planetary magnetosphere inject kinetic energy into the planetary atmosphere .
the power released by the reconnection per area , corresponding flux , is estimated as @xmath136 where @xmath137 is the magnetic permeability of the plasma , @xmath138 and @xmath139 are the magnetic field at the pole of a star and a planet .
if adopting @xmath140 g , @xmath141 g , and @xmath142 , we obtain @xmath143 , ( @xmath144 is the vacuum permeability ) .
since this value is larger than our fiducial value of @xmath145 in section [ ss : xuv ] , it is possible to account for the mass loss rate .
we performed a time - series analysis of the transit depth of kic 12557548b .
we found that the transit depth variation has three values of specific periodicity above fap=0.1 % and that the shortest one @xmath3 days is consistent with the stellar rotation period estimated by visibility modulation .
we interpreted these results as evidence that the evaporation is related to stellar activity .
using the theoretical mass - radius relation with tide , we showed that the massive escape of a rocky planet driven by xuv is possible if assuming high xuv flux and high efficiency .
the energetic electrons released by the magnetic reconnection is another possible scenario . though we could not identify any plausible scenario to account for the @xmath146/gyr evaporation correlated with stellar activity
, future xuv observations will enable us to understand what type of energy evaporates close - in small planets , if transiting planet surveys such as _ transiting exoplanet survey satellite _ find nearby similar systems .
we thank yoichi takeda for helping with the spectroscopic analysis .
this work is supported by the astrobiology project of the cnsi , nins ( ab251009 ) and by grant - in - aid for scientific research from jsps and from the mext ( nos .
25800106 , 25400224 , and 25 - 3183 ) . | violent variation of transit depths and an ingress - egress asymmetry of the transit light curve discovered in kic 12557548have been interpreted as evidences of a catastrophic evaporation of atmosphere with dust ( @xmath0 ) from a close - in small planet . to explore what drives the anomalous atmospheric escape , we perform time - series analysis of the transit depth variation of _ kepler _ archival data for @xmath1 3.5 yr . we find a @xmath2% periodic variation of the transit depth with @xmath3 days , which is within the error of the rotation period of the host star estimated using the light curve modulation , @xmath4 days .
we interpret the results as evidence that the atmospheric escape of kic 12557548bcorrelates with stellar activity .
we consider possible scenarios that account for both the mass loss rate and the correlation with stellar activity .
x - ray and ultraviolet ( xuv)-driven evaporation is possible if one accepts a relatively high xuv flux and a high efficiency for converting the input energy to the kinetic energy of the atmosphere .
star - planet magnetic interaction is another possible scenario though huge uncertainty remains for the mass loss rate . |
in many applications we know that the data or measurement vectors do not fill the whole vector space but that they are confined to some special regions .
often the form of these regions is defined by problem specific constraints .
typical examples are color vectors .
the values in these vectors represent photon counts and therefore they must be non - negative and color vectors are all constrained to lie in the positive part of the vector space . apart from the measurement vectors we are also interested in transformations acting on these measurements .
a typical example in the case of color is the multiplication of the vectors with a global constant .
this can describe a global change of the intensity of the illumination or a simple change in the measurement units .
it can be shown that many non - negative measurement spaces can be analyzed with the help of lorentz groups acting on conical spaces .
details of the mathematical background and investigations of spectral color databases can be found in @xcite .
similar geometric structures are used in perception - based color spaces such as the standard cielab color system . here the l - part ( related to intensity ) is given by the positive half - axis .
the ( a , b)-part describing the chromatic properties of color are often parameterized by polar coordinates where the radial part corresponds to saturation ( given by a finite interval ) and the angular part represents hue .
a second example that will be considered comes from low - level signal processing .
basic edge - detectors for two - dimensional images come in pairs and the resulting two - dimensional result vectors are often transformed into polar coordinates where the radial part describes the edge - magnitude and the angular part the edge orientation . in many machine
learning approaches the raw magnitude value is further processed by an activation function which maps the non - negative raw output to a finite value .
a typical activation function is the hyperbolic tangent function @xmath0 and we see that output of these edge detector systems are located on the unit disk . in the following
we describe how harmonic analysis ( the generalization of fourier analysis to groups ) can be used to investigate functions defined on special domains with a group theoretical structure .
we will present the following results : the investigated functions are defined on a domain with a group of transformations acting on this domain and in this case they can be interpreted as functions on the transformation group .
the transformation group defines an ( essentially unique ) transform that shares many essential properties of the ordinary fourier transform .
examples of such properties are the simple behaviour under convolutions , preservation of scalar products in the orignal and the transform domain and the close relation to the laplace operator .
this will be illustrated for the unit disk and its transformation group @xmath1 .
the functions corresponding to the complex exponentials are in this case the associated legendre functions and the ( radial ) fourier transform on the unit disk is the mehler - fock transform ( mft ) .
the associated legendre functions are also eigenfunctions of the hyperbolic laplace operator and the mft can therefore be used to study properties related to the heat equation on the unit disk . as mentioned before we will illustrate the theory with some examples from color processing and low - level signal processing .
we will mainly study probability distributions on the unit disk and show how the mft can be used to investigate textures and to manipulate color properties of images .
probability distributions on manifolds and their kernel density estimators have been studied before ( see , for example , @xcite , @xcite,@xcite , @xcite @xcite and @xcite ) .
often they are assuming that the manifold is compact . here
we study functions on the unit disk as the simplest example of a non - compact manifold but apart from the riemannian geometry it also has a related group structure which is crucial in the derivation of the method .
the mft is a well - known tool in mathematics and theoretical physics but there are very few studies relevant for signal processing .
some examples are described in @xcite and @xcite describes an application of hyperbolic geometry in radar processing .
an application to inverse problems is described in @xcite . to our knowledge
this is the first application to color signal processing .
the structure of the paper is as follows : first we introduce the relevant group @xmath1 and describe how it operates on the unit disk .
we also collect some basic facts of the disk - version of hyperbolic geometry and introduce the parametrization of the unit disk by elements of the factor group @xmath2 .
next we introduce some facts from the representation theory of the group @xmath1 and its relation to the associated legendre functions .
restriction to zonal functions ( that are functions of the radius only ) leads to the mft which is a central result in the harmonic analysis of this group .
the theoretical concepts will be illustrated with two examples , one from color science and the other from texture analysis .
for the color image we show how the mft can be used to desaturate local color distributions and for the textures images we show how it can be used to characterize the textures in the normalized brodatz texture database described in @xcite .
no claim is made regarding the performance of the method compared with other methods in color or texture processing .
in the following we consider points on the unit disk as complex variables @xmath3 and we introduce the group @xmath1 consisting of the @xmath4 matrices with complex elements : @xmath5 the group operation is the usual matrix multiplication and the group acts as a transformation group on the open unit disk @xmath6 ( consisting of all points @xmath7 with @xmath8 ) as the mbius transforms : @xmath9 the concatenation of two transforms correspond to the multiplication of the two corresponding matrices which gives : @xmath10 for all matrices and all points . in the following
we will use the notation @xmath11 if we think of the group elements as matrices .
if we want to stress their role as elements of a group or if we want to parametrize the mbius transforms then we will often use the symbol @xmath12 .
every three - dimensional rotation can be written as a product of three rotations around the coordinate axes .
a similar decomposition holds for @xmath1 whose elements are : @xmath13 with three parameters @xmath14 defining subgroups @xmath15 and @xmath16 we write the decomposition as @xmath17 this is known as the cartan or polar decomposition of the group and @xmath18 are the cartan coordinates .
matrices in @xmath19 are rotations , leaving the origin fixed and applying a general element in @xmath1 to the origin @xmath20 gives : @xmath21 defining two transformations as equivalent if the parameters @xmath22 are identical , we obtain a correspondence between the points on the unit disk and the equivalence classes of transformations .
we write @xmath23 and functions on the unit disk are functions on @xmath1 that are independent of the last argument of the cartan decomposition . for two points @xmath24
we define the kernel function @xmath25 as the probability of obtaining the true measurement @xmath26 given that the actual measurement resulted in the point @xmath3 .
often it is natural to require @xmath27 for example in the case of invariance against coordinate system changes .
the mbius transforms define an invariant , hyperbolic , distance between points @xmath3 and @xmath28 given by @xmath29 and we consider only kernel functions of the form @xmath30 .
points on the unit disk are equivalence classes of group elements and in the general case we consider ( ( left - isotropic ) kernels @xmath31 on the full group satisfying @xmath32 it follows that @xmath33 where @xmath34 is the identity element . in the case of the unit disk
we consider elements @xmath35 and @xmath36 and we find the relation between these two decompositions and the difference @xmath37 as ( see @xcite , ( vol 1 , page 271 ) ) : @xmath38 we therefore consider only radial kernels of the form : @xmath39 in general we have to compute the function values @xmath40 for every pair @xmath41 of special interest are functions which separate these factors .
they are the associated legendre functions ( zonal or mehler functions , @xcite , page 324 ) of order @xmath42 and degree @xmath43 defined as ( see eq.([eq - alfundef ] ) ) : @xmath44 with the addition formula ( @xcite , page 327 ) @xmath45 where @xmath46 .
it seperates dependent terms , ( from data @xmath47 with parameters @xmath48 ) and the the kernel part ( from @xmath49 with @xmath50 ) .
this is generalized by the mft , showing that a large class of functions are combinations of the associated legendre functions : [ the : mft ] for a function @xmath51 defined on the interval @xmath52 define its transform @xmath53 as : @xmath54 then @xmath51 can be recovered by the inverse transform : @xmath55 details about the transform , special cases and its applications can be found in @xcite,@xcite,@xcite,@xcite and @xcite . in the general group theoretical context the integral in eq.([mft - cos ] )
defines a convolution over the subgroup @xmath56 parametrized by the hyperbolic angle @xmath57 in the case of the group @xmath58 this general construction can be applied by using the cartan coordinates of group elements .
the analysis of the angular parts lead to ordinary fourier series and for the radial part we find for the mft ( in hyperbolic coordinates ) the integrals ( using eq.([eq - ttrans ] ) characterizing the contribution from datapoint @xmath59 : @xmath60 with @xmath61 .
the coefficients @xmath62 are computed from the measurements and the weights @xmath63 are independent of the data .
it is also known that the mft preserves the scalar product ( parseval relation ) . if @xmath64 then @xmath65 this can be found in @xcite(7.6.16 ) and @xcite(7.7.1 ) where the interested reader can find more information on the mft .
finally we mention that the conical functions are eigenfunctions of the laplace - operator which is the operator that commutes with the group action .
the eigenvalues are @xmath66 ( see @xcite chap . 4 for a brief summary of these results ) .
in the first example we use the mft to study a simple filter system ( more information can be found in @xcite ) .
the input of the system are the pixel values on a four - point orbit of the dihedral group , for example the four corner points of a square .
the theory of group representations for the dihedral group shows that a natural filter pair consists of the edge filters with coefficients @xmath67 and the raw output values @xmath68 should be transformed into polar coordinates @xmath69 they are the familiar edge magnitude ( independent of dihedral transformation ) and the edge - orientation . following popular practice in machine learning the raw magnitude result @xmath70
is converted to a value @xmath71 less than or equal to one with the help of a transfer function .
every configuration of a four - pixel distribution results thus in a point on the unit disk . in our first experiment
we compute the probability distribution of such a filter system for each of the 112 images in the normalized brodatz texture database ( described in @xcite ) .
the kernels used for the density estimators are of the form @xmath72 for which the mft is known .
the resulting probability density functions on the disc where then analysed with the help of the mft and the ordinary fourier transform in the angular variable . in the following example we used the orientation invariant component of the angular fft and
then computed its mft - values .
next we applied the parseval relation ( eq.([eq : parseval ] ) ) to compute the distance between pairs of mfts .
multidimensional scaling was then used to project the distance distributions to the two - dimensional plane . ignoring the orientation information and only using the magnitude values results in a characterization of the textures in terms of overall roughness which is illustrated in figure [ fig : brodatz ] .
it turns out that the first axis in the multidimensional scaling mapping is by far the most dominant and we therefore sort the textures using their position on this first axis .
the contour plots of the sorted sequence of these pdfs is shown in figure [ fig : brodatz](a ) .
the first and last five textures in the sequence are shown in figure [ fig : brodatz](b ) .
\(a ) sorted densities \(b ) first and last textures in the sequence in the color image illustration we convert first the original rgb image into its cielab form .
the ( a , b ) part of a cielab vector describes the chroma properties of a color and we convert the original ( a , b ) vector to polar coordinates and scale the radial part so that all points are located on the unit disk . for each point in the image
we then select a window ( of size @xmath73 pixels ) and compute the mft of the distribution of the radial ( saturation ) variable . from the fact that the conical functions are eigenfunctions of the laplace operator it follows that we can simulate a kind of heat - flow on these probability distributions of the saturation values . in the mft space
the `` high - frequency '' components with big @xmath74 values are faster suppressed due to the negative eigenvalue of the conical functions . in the experiment we computed first the probability distributions of the local saturation values .
we then used the mft to transform these original distributions and then we used a sequence of increasing time - intervals to simulate the dissipation of the distributions . due to the form of the conical functions we expect that image points with concentrated high saturation values should survive longest whereas low saturation values and inhomogenous distributions should disappear faster .
after the operation in the transform domain we transformed the new function back and we detected the mode ( saturation with the highest probability value ) .
this mode - saturation value is used as the new saturation at this point , the angular - hue and the intensity value of the cielab l - component are copied from the original and the color vectors are transformed back into rgb - form .
the result of a sequence of sixteen such time - intervals is shown in figure [ fig : pepmon ] . here
the first fifteen use equal time - increments whereas the last image corresponds to a time - interval that is twice as long as the previous one .
we observed that many processes produce signals that are located on the unit disk .
apart from the property that the outpout vectors are points on the unit disk it is also important that operations on the disk have a meaningful interpretation . in the case of the edge filter systems these operations are changes in orientation and edge - magnitude . for the chroma - vectors the corresponding operations are hue and saturation changes .
the next step is the observation that functions on the unit disk have an interpretation as functions on the lorentz group @xmath75 for this group there is a ( more or less ) unique transform with similar properties as the ordinary fourier transform .
this transform is the mehler - fock transform and the functions that correspond to the complex exponentials are the conical functions .
based on the properties of the mft we showed that ( a ) processing can be done in the original or the transform domain ( b ) the relation between the mft and group convolution leads to a separation of the kernel density estimator into data - dependent and kernel - dependent terms ( c ) the mft preserves geometry between the original and the transform domain and we can therefore estimate similarity between signals in either space and ( d ) we can use the eigenfunction property of the laplacian to study methods for smoothing the data . the detailed study of numerical aspects such as approximation errors , sampling properties , and comparison to other methods is left for future investigations .
e.a chevallier , f.b barbaresco , and j.a .
angulo , `` probability density estimation on the hyperbolic space applied to radar processing , '' in _
geometric science of information_. 2015 , vol .
9389 of _ lncs _ , pp . 753761 , springer verlag . | many stochastic processes are defined on special geometrical objects like spheres and cones .
we describe how tools from harmonic analysis , i.e. fourier analysis on groups , can be used to investigate probability density functions ( pdfs ) on groups and homogeneous spaces .
we consider the special case of the lorentz group su(1,1 ) and the unit disk with its hyperbolic geometry , but the procedure can be generalized to a much wider class of lie - groups .
we mainly concentrate on the mehler - fock transform which is the radial part of the fourier transform on the disk .
some of the characteristic features of this transform are the relation to group - convolutions , the isometry between signal and transform space , the relation to the laplace - beltrami operator and the relation to group representation theory .
we will give an overview over these properties and their applications in signal processing .
we will illustrate the theory with two examples from low - level vision and color image processing . |
in the companion paper @xcite we compare the force networks in tapped systems by using relatively simple measures : probability density functions ( pdfs ) for normal and tangential forces between the particles , correlation functions describing positional order of the considered particles , as well as possible correlations of the emerging force networks .
these classical measures are supplemented by analysis of cluster sizes and distributions at different force levels ( i.e. , by considering the part of the force network that only includes contacts involving forces exceeding a threshold ) .
these results have uncovered some differences between the force networks in the considered systems .
for example , we have found that the number of clusters as a function of the force level is heterogeneous in the tapped systems under gravity , with different distributions deeper in the samples compared to the ones measured closer to the surface .
however , some of the differences remain unclear .
for example , tapped disks exposed to different tap intensities that lead to the same ( average ) packing fraction , are found to have similar pdfs and similar cluster size distributions , although it is known @xcite that there are some differences in the geometrical properties of the contact networks in these systems . in the present paper ,
we focus on a different approach , based on persistence analysis .
this approach has been successfully used to explain and quantify the properties of force networks in the systems exposed to compression @xcite .
in essence , persistence analysis allows to quantify the force network ` landscapes ' in a manner that is global in character , but it still includes detailed information about the geometry at all force levels . the global approach to the analysis of force networks makes it complementary to other works that have considered in detail the local structure of force networks @xcite , and attention to geometry distinguishes this approach from network type of analysis @xcite .
we will use persistence analysis to compare the force networks between the systems of disks exposed to different tapping intensities , as well as to discuss similarities and differences between the systems of disks and pentagons . as we will see , some differences between the considered networks that could not be clearly observed ( and even less quantified ) using classical measures become obvious when persistence analysis is used .
furthermore , persistence analysis allows for formulating measures that can be used to quantify , in a precise manner , differences in force networks between realizations of a nominally same system .
we note that persistence has been used to quantify the features in other physical systems such as isotropically compressed granular media @xcite .
it was also used to study dynamics of the kolmogorov flow and rayleigh - bnard convection @xcite .
this paper is organized as follows . in sec .
[ sec : methods ] , we discuss briefly the persistence approach and also provide some examples to illustrate its use in the present context . in sec .
[ sec : results ] , we discuss the outcome of persistence approach and quantify the differences between the considered systems .
section [ sec : conclusions ] is devoted to the conclusions and suggestions for the future work .
the simulations utilized in this paper are described in detail in @xcite ; here we provide a brief overview .
we consider tapped systems of disks and pentagons in a gravitational field .
the particles are confined in two - dimensional ( 2d ) rectangular box with solid ( frictionless ) side walls , .
initially @xmath0 particles are placed at random ( without overlaps ) into the box , and the particles are allowed to settle to create the initial packing .
then , @xmath1 vertical taps are applied to each system considered ; we discard the initial @xmath2 taps and analyze the remaining @xmath0 . after each tap
, we wait for the particles to dissipate their kinetic energy and achieve a mechanical equilibrium .
we record the particles positions and the forces acting between them ; the interactions between the particles and the walls are not included . for more direct comparison ,
the forces are normalized by the average contact force .
in addition to discussing the influence of particle shape , we consider two different tapping intensities , @xmath3 ( called ` high ' and ` low ' tap in what follows ) that lead to the same packing fractions for disks ( @xmath4 ( low ) and @xmath5 ( high ) , where @xmath6 is the disk radius and @xmath7 the acceleration of gravity ) .
we also discuss the influence of gravitational compaction , and for this purpose we consider ` slices ' of the systems , @xmath8 particle diameters thick : bottom slice positioned deep inside the domain , and the top slice close to the surface .
see @xcite for more details .
we are interested in understanding the geometry exhibited by force networks .
their complete numerical representation contains far too much information . with this in mind
, we make use of the tools from algebraic topology , in particular homology , to reduce this information by counting simple geometric structures . in the two dimensional setting of interest in the present context , fixing a magnitude , @xmath9 , of the force and considering the particles which interact with a force at or above @xmath9 yields a 2d topological space , @xmath10 .
two simple geometric properties of @xmath10 are the number of components ( clusters ) , @xmath11 , and the number of loops ( holes ) , @xmath12 . in @xcite it is shown that even though we are counting very simple geometric objects , by varying the threshold @xmath9 , the set of betti numbers @xmath11 and @xmath12 provides novel distinctions between the behavior of the above mentioned systems .
however , there is an obvious limitation to just using the betti numbers to describe a system .
consider two different thresholds and assume that the values of the betti numbers are the same .
does this mean that the geometric structures , e.g. components and loops , are the same at these two thresholds , or have some components or loops disappeared and been replaced by an equal number of different components or loops ?
this distinction can not be determined from the betti number count alone . to provide more complete description , we make use of a relatively new algebraic topological tool called _ persistent homology_. in the context of the 2d systems that we are considering here , it is sufficient to remark that to each force network landscape persistent homology assigns two _ persistence diagrams _ , @xmath13 and @xmath14 , such as those shown in fig .
[ fig : diagrams ] .
each persistence diagram consists of a collection of pairs of points @xmath15 where @xmath16 , the birth , indicates the threshold value at which a geometric structure ( a component / cluster for @xmath13 or a loop for @xmath14 ) first appears and @xmath6 , the death , indicates the threshold value at which the geometric structure disappears . in this paper
we measure the geometry of the part of the contact network with force interactions greater than a given threshold , and thus @xmath17 .
the value @xmath18 is called the _ lifespan_. note that the component represented by the point @xmath19 ` dies ' when it merges with some other component with the birth coordinate larger or equal to @xmath16 .
in particular , the single generator in @xmath13 with death coordinate @xmath20 , see fig .
[ fig : diagrams](a ) , represents the component that contains the strongest force ` chain ' in the system : the one that formed at the highest force level .
note that it has both the highest birth value and the longest lifespan .
more detailed interpretation of @xmath13 in 1d can be found in @xcite , while a rigorous presentation for 2d is given in @xcite .
+ the loop structure of a force network is described by @xmath14 .
a loop in the network is a closed path of the edges connecting centers of the particles .
similarly to @xmath13 , the point @xmath21 indicates that a loop appears in the part of the network exceeding the force threshold @xmath16 .
this loop is present for all the values of the threshold in @xmath22 $ ] . at the value
@xmath6 , this loop is filled in , that is , the interior of the loop is filled in with particles that form a crystalline structure , and the forces between the interacting particles inside of the loop are larger or equal to @xmath6 .
this fill - in process can be also seen as filling the loop by ` trivial ' loops formed by exactly three particles with forces stronger or equal to @xmath6 . to visually distinguish the loops that do not get filled in at any force level ( including the threshold @xmath23 )
, we set their death time to @xmath24 .
however , in the distance computations ( discussed below ) we follow the convention presented in @xcite and replace them by @xmath23 .
note that in the example shown in fig .
[ fig : diagrams ] , for pentagons there is a single loop that gets filled in ( close to threshold level @xmath23 ) , while for disks there is a number of loops that gets filled in at a variety of thresholds .
a particularly simple descriptor based on the lifespans is the total persistence , @xmath25 i.e. the sum of all lifespans of the points in @xmath26 . in the context of fig .
[ fig : diagrams ] , the total persistence of ( a ) and ( c ) is roughly the same , while the total persistence of ( b ) is roughly twice that of ( d ) .
thus , this gives a simple measure that can , at least in some settings , distinguish persistence diagrams arising from different systems . to further distinguish the diagrams , we will also consider the distribution of lifespans ( representing the number of times a value of lifespan in a specified range is found ) . as is made clear in section [ sec : results ] , even within a single system there can be considerable variability in the force networks from tap to tap .
unfortunately , the concept of an average persistence diagram is not yet well defined @xcite .
however , using the above mentioned measures we avoid this difficulty by applying them to an aggregate persistence diagram obtained by considering , as is done in fig .
[ fig : diagrams_all ] , all persistence points from the @xmath0 simulated taps on a single diagram .
the distributions of birth times presented in section [ sec : results ] are based on these diagrams .
note that for the remainder of the paper the distributions of birth times and lifespans are _ normalized by the number of particles _ in the domain used to define the force network under consideration .
we note that the @xmath26s provide a compressed and simplified description of the underlying force network landscape .
thus , some information , such as size , shape and position of the components or loops , is discarded while passing from a force network to the corresponding @xmath26 . therefore , different force networks , may produce the same @xmath26s . however , as is discussed in sec .
[ sec : distance ] , there are metrics that can be imposed on the space of all persistence diagrams such that if two force network landscapes are similar , then their associated persistence diagrams are similar . on the other hand ,
if the diagrams differ considerably , then so do the corresponding force networks .
hence , in summary , persistent homology provides a continuous reduction of information that captures geometric information .
it is worth mentioning that persistence computations extend to higher dimensions . in particular , for @xmath27d , the diagram @xmath13 describes the structure of the connected components , as in @xmath28d .
the features in @xmath14 are interpreted as tunnels rather than loops .
finally , there is an additional diagram @xmath29 that describes the structure of cavities .
the computational codes used to construct the force networks and persistence diagrams are available at @xcite , respectively .
there are also other publicly available packages for computing persistent homology @xcite .
in the previous section , we introduced specific descriptors of persistence diagrams based on a single feature of the points , e.g. lifespan or birth value . in this section
we discuss metrics on the space of persistence diagrams that are based on the entire diagram , i.e. we compare two diagrams by comparing all points in each diagram . note that this comparison does not involve force thresholding : this measure compares the force networks at all force levels .
consider two persistence points @xmath30 and @xmath31 .
the distance between @xmath32 and @xmath33 is defined by @xmath34 now , given two persistence diagrams @xmath26 and @xmath35 let @xmath36 be a bijection between points in the two persistence diagrams where we are allowed to match points of one diagram with points on the diagonal of the other diagram .
degree-@xmath37 wasserstein distance _ , @xmath38 ,
is obtained by considering for each bijection , @xmath39 , the quantity @xmath40 and defining the distance between @xmath26 and @xmath35 to be the minimum value of this quantity over all possible bijections . stated formally @xmath41 the _ bottleneck distance _
@xmath42 is given by @xmath43 the cost of ` moving the points ' ( i.e. , selecting a given bijection ) varies for different distances .
the bottleneck distance captures only the largest move corresponding to the largest difference between the diagrams . on the other hand , the wasserstein distance @xmath44 sums up all the differences with equal weight .
if all the points in one diagram are close to the points in the other diagram ( or close to the diagonal ) , then the bottleneck distance is small .
however , @xmath44 tends to be large since it is a sum over a large number of small differences . in the case that @xmath44 is close to @xmath45 the diagrams tend to differ in a small number of points .
there is the following relation between the distances : @xmath46 for @xmath47 and @xmath48 converges to @xmath45 as @xmath49 goes to infinity .
hence , using the @xmath50 distance keeps track of all the changes but the small differences contribute less . comparing the @xmath51 and the @xmath50 distances allows to better understand the difference between the diagrams .
for example if @xmath45 and @xmath50 are similar but @xmath44 is large , then there is only one dominant difference between the diagrams and a large number of small differences . to compare a large number of @xmath26s representing the steady states of the tapped systems , we use the distance matrix ( heat map ) @xmath52 .
the entry @xmath53 is the distance between the diagrams corresponding to the taps @xmath54 and @xmath55 . clearly , @xmath56 and
if the states are similar , then the value @xmath53 is small .
the distance matrix provides a detailed information about the differences between all the states .
sometimes a more condensed representation of the differences between the states is desirable ; for this purpose we will use distribution of the values of @xmath52 . to illustrate that the distance matrices are sensitive to the structure of the force networks , and furthermore that they allow for a simple visual inspection of the similarities and differences both between the considered systems and between different realizations / taps for the same system , we provide here an example .
we consider a system of disks exposed to high intensity tapping , record all the force information , and then randomly mix up the forces .
this randomization is done after each tap by performing @xmath57 force swaps ( i.e. , picking any two contacts at random and swapping the forces between them ) .
this procedure leaves the pdfs of the force networks unchanged , however the force network may undergo a dramatic change , since one expects the force chains to be broken into a large number of short chains and therefore the differences between the states should be larger .
figure [ fig : mixed ] shows the @xmath44 distance matrix for the original and randomized system .
considering the distances between @xmath13 diagrams first , we see that the distances between ` real ' realizations are small compared to the ones between randomized systems , and also that the distances between the original and randomized states are larger than the ones between the randomized states . on the contrary ,
the variation in @xmath14 diagrams is smaller for the randomized systems .
this is due to the fact that by randomly reassigning the forces , the loops with strong force interactions are destroyed and the points in @xmath14 tend to be close to the diagonal .
we conclude that the distance matrices clearly identify the differences between original simulation results and the randomized ones .
as we will see in what follows , they can be also used to identify the differences between various simulation results considered in the present work .
gravitational compaction and its influence on force networks is discussed in @xcite , where it is shown that pdfs of the normalized forces do not depend on the depth in the sample .
however , using @xmath58 to count components shows differences : for both normal and tangential forces the number of components for the bottom slices is considerably larger than for the top ones . in this paper
we compare top and bottom slices using the corresponding distance matrices and the descriptors discussed in the previous section .
+ + + figures [ fig : disks_low_heat ] shows the @xmath44 distance matrices for the persistence diagrams of top and bottom slices of disks packings , respectively .
while distances between the components do not appear to vary significantly , three observations can be made regarding @xmath14 distances , shown in the parts ( c ) and ( d ) : ( 1 ) the distance between persistence diagrams of the top slice associated with different taps is relatively small ( see lower right corners ) ; ( 2 ) persistence diagrams of the bottom slice exhibit more variability ( see upper left corners ) , and ( 3 ) the distances between persistence diagrams for different taps from the top slice and bottom slice range from small to large ( see upper right corners ) . to reinforce these observations , we include fig . [
fig : disks_low_hist ] that contains plots of the distribution of the distances .
the distributions are computed only from the upper triangle of the distance matrix .
hence the diagonal ( zero ) entries are not included .
one immediate conclusion from the observation ( 3 ) is that the geometry as measured by persistence diagrams for the force network observed after each tap for the top slice are not far away from those for the bottom slice , but lay on a small subdomain of this second more scattered set .
hence , the cross comparison distances are dominated by the scattered distances of the bottom slice networks .
these distinctions are much less pronounced in fig .
[ fig : disks_low_heat_b ] , that shows @xmath45 distance .
this suggests that there is no single pronounced difference in the geometry of the force networks .
in addition , @xmath50 ( picture not shown for brevity ) , turns out to be quantitatively similar to the @xmath45 distance matrices . based on the discussion in sec .
[ sec : distance ] , this fact indicates that the force networks for the bottom and top slices are similar , but the number of small variations in the birth and death of each feature ( cluster or loop ) from tap to tap are more prominent for the bottom slice . in particular , the light and dark bands of fig .
[ fig : disks_low_heat](c ) and ( d ) suggest the existence of further structure in our systems that we explore in the next section
. the measures considered in @xcite ( force pdfs , betti numbers ) do not identify differences in the properties of the force networks in the systems of disks exposed to different tapping intensities that lead to the same packing fraction .
however , earlier work @xcite , based on different type of simulations where the side walls were frictional , which possibly influences the force network structure , suggested that some ( rather difficult to observe ) differences may exist . in this section we show that analysis based on persistent homology provide additional insight that allows to understand the origin of the differences , both when considering averaged results , and on the level of individual realizations and their variability .
careful inspection of the distance matrices for different distance definitions , normal and tangential forces , components and loops , uncovers the following facts .
first , we do not observe any appreciable differences between the distances when components / clusters are considered ( figures not shown for brevity ) .
therefore , we expect that the distribution of components between the disk - based systems exposed to high and low tap intensities are similar ( however see below )
. this conclusion does not apply to the distances between @xmath14 diagrams .
figure [ fig : disks_bottom_heat_hist](a , b ) shows the corresponding results , for @xmath59 distance .
there is a clear difference between high and low tapping regimes , suggesting that the structure of the loops is very different .
figure [ fig : disks_bottom_heat_hist](c , d ) shows the corresponding distributions confirming a clear separation between the two cases considered . in comparison to low tapping states , high tapping
is characterized by force networks that are significantly more similar to each other , both for normal and tangential forces .
we emphasize that this difference is difficult to observe and quantify by using any other measure we are aware of .
we also observe in fig .
[ fig : disks_bottom_heat_hist](c , d ) that the distances between the realizations for low tapping are as large as the distances between low and high tapping regimes ; we will discuss this finding in more detail in what follows .
+ we now proceed to answer the following two questions : ( i ) what is the difference between the structure of the networks for high and low tapping ?
( ii ) figure [ fig : disks_bottom_heat_hist](a , b ) shows that there appears to be some structure in the evolution of the distances for @xmath14 persistence diagrams .
what is the origin of this structure ? to answer the first question , we now consider other measures that can be computed from @xmath60s , starting with birth times .
figure [ fig : disks_birth ] shows the distribution of birth coordinates for the considered @xmath13 and @xmath14 persistence diagrams . as a reminder , birth time indicates at what force threshold level the considered features ( components , loops ) appear .
perhaps surprisingly , we see that not only the structure of loops is different , but the structure of components differs as well : for low tapping regime , there is a much more pronounced peak , particularly for the normal forces . while consistent difference was seen when @xmath58 was considered ( see fig .
9 in @xcite ) , the difference is more pronounced when considering birth times .
figure [ fig : disks_birth ] shows that in general there are more features born at every force level for low tapping regime . in principle , it could happen that the lifespans of these extra features are relatively small and the systems only differ in minor fluctuations .
distributions of lifespans , shown in fig . [ fig : lifespanshilow ] , demonstrate that this is not the case .
the distributions of lifespans for low tapping regime are typically above the distributions for the high tapping regime .
therefore , even the number of prominent features is typically larger for the low tapping regime .
however , there is an important exception . for normal forces
the number of components with lifespan larger than one is larger for the high tapping regime .
the crossover of the @xmath13 distributions for the normal force around the lifespan equal one leads to the fact that the corresponding distributions of total persistence , shown in fig .
[ fig : totalpers - hilow](a ) , are centered at the same point . as expected , the remaining distributions for the low tapping regime , shown in fig .
[ fig : totalpers - hilow ] , are shifted to the left .
+ while the interpretation of the results , such as the differences of birth times and lifespans may not be immediately obvious , these measures clearly show the ability of the persistence analysis to distinguish between the considered systems . on a practical side
, the fact that the differences _ can _ be identified suggests that the considered systems have different macroscopic properties .
+ + so far , we have discussed one of the questions listed above , regarding the differences between the considered systems .
now we proceed to discuss the second one related to the origin of the structure apparent in fig .
[ fig : disks_bottom_heat_hist ] for the low tapping regime .
recall ( see ( * ? ? ?
iii b ) ) that low tap intensity often does not lead to significant changes of the packings of disks from one tap to the next .
therefore , one may expect that for a certain number of taps , the realizations are correlated , close to each other ( with small distances between them ) , until the configuration of the particles , and the corresponding force network , change dramatically . the question is whether these ` jumps ' from one set of similar packings to another one are captured by the @xmath60s . to show that the answer to this question is positive , we recall that the packing fraction @xmath61 may change from one tap to the next .
since @xmath61 values obtained for different taps provide a rather noisy signal , we consider instead its auto - correlation function , as well as instantaneous cross - correlation between @xmath61 and total persistence , @xmath62 ( for simplicity , we consider here @xmath62 instead of the distances ) .
regarding total persistence , we focus on @xmath63 .
figure [ fig : correlation ] gives the results for the auto - correlation functions defined for any descriptor @xmath64 as @xmath65 $ ] , where @xmath66 indicates an average over all time origins , and @xmath67 is used to normalize so that @xmath68 . auto - correlation curves for @xmath61
show that while for high tapping there is no observable correlation between taps , for low tapping there is a clear correlation for up to @xmath69 taps .
this correlation is consistent with the structure of the distance matrices for low tapping , see fig .
[ fig : disks_bottom_heat_hist](a , b ) .
the results for @xmath63 auto - correlation functions are very similar .
we can go further and compute the instantaneous cross - correlation @xmath70 between @xmath63 and @xmath61 , defined as @xmath71 [ tp({{\mathsf{pd}}}_1)(t)-\langle tp({{\mathsf{pd}}}_1 ) \rangle ] \rangle}{\sigma_{\phi}\sigma_{tp({{\mathsf{pd}}}_1)}},\ ] ] where @xmath72 indicates the variance of the variable .
we find significant correlation between these two quantities : @xmath73 reaches the values of @xmath74 for normal and tangential forces ( here @xmath75 means perfect correlation and @xmath23 lack of correlation ) . to rationalize the correlation of @xmath61 and @xmath63 ,
we recall the results obtained by considering the systems of disks exposed to compression @xcite .
in that system it was found that for monodisperse disks , which are more likely to crystallize ( therefore having larger @xmath61 ) , there is a larger number of points in @xmath14 , consistently with the results presented here .
we note that in @xcite it was found that larger number of points in @xmath14 occurred for strong forces ( with the idea that strong loops form at the boundaries of the fault zones separating crystalline regions ) ; we expect a similar reason for the larger values of @xmath63 in the present setting .
to conclude this section , we note that persistence analysis allows to clearly identify differences between the systems of disks exposed to different tapping intensities leading to the same ( average ) packing fraction : these differences are particularly clear when considering the structure of loops .
the differences are apparent for the averaged persistence diagrams but they are even more prominent when considering individual taps and their variability .
this variability is much stronger for the systems exposed to low tapping . as already noted in the context of the results shown in fig .
[ fig : disks_bottom_heat_hist](c , d ) the differences between different realizations for low tapping may be as large as the differences between low and high tapping ones .
in @xcite we discussed some of the differences in the structure of the force networks between disks and pentagons .
the main findings reported in that paper are that the differences between these systems manifest themselves particularly in the structure of tangential force networks measured by @xmath58 ( although pdfs of the forces are almost indistinguishable ) , and by the number of loops , measured by @xmath76 , for both normal and tangential forces .
the number of loops in the disk - based system is consistently larger .
this finding supports the idea that the clusters are larger for disks , and therefore can support larger number of loops . in the present work
, we will discuss additional insight that can be reached by persistence analysis .
figures [ fig : disks_pents_heat_w1 ] and [ fig : disks_pents_hist_w1 ] show the distance matrices and corresponding distributions comparing disks and pentagons exposed to the same ( low ) tapping intensity . in agreement with the results from @xcite , the differences between the components ( the parts ( a ) and ( b ) of these figures ) are relatively minor .
considering loops , these figures show that the distances between pentagon - based systems are much smaller than for the disk - based ones .
in particular , fig .
[ fig : disks_pents_hist_w1 ] shows that the distances between pentagon - based systems are centered at much smaller values , and their distribution is much narrower than for disks .
we also note that the distances between disks and pentagons are much larger than between different disk realizations , showing that persistence analysis clearly distinguishes these systems .
we note that consideration of other distances , such as bottleneck , @xmath45 , that measures only the largest difference , are consistent with the ones presented for @xmath44 distance ( figures not shown for brevity ) .
in particular , the distributions of @xmath45 for loops are similar to the ones shown in fig .
[ fig : disks_pents_hist_w1 ] , with the maximum and the spread for pentagons smaller than for disks .
this is as expected , since the loops form at lower force level in pentagon - based system , compare fig . [
fig : diagrams_all](c ) and ( d ) .
+ + we proceed by discussing the source of the differences between the disk- and pentagon - based systems considered so far .
first , we focus on the distributions of birth times
. figsures [ fig : disks_pents_birth0.1 ] and [ fig : disks_pents_birth0 ] show the corresponding results .
the only difference between these figures is that in fig .
[ fig : disks_pents_birth0.1 ] we consider only the points with the lifespan larger than @xmath77 , while in fig .
[ fig : disks_pents_birth0 ] we consider all the points .
the reason for showing both figures is that the differences between the two provide additional information about the points with short lifespan .
considering components for normal forces , parts ( a ) in these two figures , we observe that birth times capture some differences between the two systems that were not obvious when considering distances .
there are more points in @xmath26s for disks that are born around @xmath78 , and more points in @xmath26s for pentagons born at larger forces .
this is consistent with the pdfs for disks and pentagons shown in fig .
10 of @xcite . for the tangential forces , parts ( b )
, we do not see much if any difference in the birth times .
regarding loops , the parts ( c ) and ( d ) of figs .
[ fig : disks_pents_birth0.1 ] and [ fig : disks_pents_birth0 ] , one consistent observation is that there are more points in @xmath26s for disks than for pentagons for the whole range of forces considered .
moreover , for disks loops start appearing at higher force level than for pentagons .
the differences between these two figures show how many of the points have a short lifespan ; these differences are particularly interesting for loops , parts ( c ) and ( d ) : we note a significantly larger number of points for pentagons at small birth times , suggesting that loops for pentagon - based systems form at very small or vanishing force , consistently with the discussion in @xcite .
this finding holds both for loops formed by normal and tangential forces .
+ figure [ fig : lifespans ] presents distributions of the lifespans for disks and pentagons . from @xmath13 diagrams we conclude that for both disks and pentagons , the dominant number of components is characterized by rather short lifespans .
we also observe a cross - over ( more pronounced for tangential forces ) between disk and pentagon distributions , although the difference is not large .
we note that the lifespans larger than @xmath79 are more probable for pentagons than for disks .
therefore , the components live longer for pentagon - based system in particular when tangential forces are considered . to use the landscape analogy , this result says that mountain peaks in the tangential force network are more pronounced for pentagon - based systems .
observe from fig .
[ fig : lifespans](c , d ) that the lifespan curves are similar to the birth time curves shown in fig .
[ fig : disks_pents_birth0 ] .
this is because for both disks and pentagons , most of the loops disappear very close to the zero force level , and thus the death time provides no additional information .
figure [ fig : totalpers ] shows the total persistence , @xmath62 , that to a large degree summarizes many of the findings discussed so far .
we recall that @xmath62 corresponds to the sum of the lifespans , see sec [ sec : methods ] , so considering the results shown in this figure together with the ones shown in fig .
[ fig : lifespans ] is useful . for @xmath80 diagrams , there is only a minor difference between disks and pentagons in the normal force network ; however , for tangential forces , there are significant differences .
this reflects the larger lifespans of the components for pentagon - based system . for @xmath63 ,
the differences are very obvious for both normal and tangential forces , and in contrast to @xmath80 results , here we find that the distribution of @xmath63 is shifted to larger values and is much broader for disk - based systems .
figure [ fig : totalpers ] shows clearly significant differences in the structure of force networks in the systems of tapped disks and pentagons .
pentagon systems tend to form new components ( clusters ) at higher force levels and these endure longer before they merge , in comparison to disk - based ones .
this is particularly evident for the tangential force network .
in contrast , loops are formed at relatively low force levels in pentagon - based systems .
hence , one could expect that the clusters that form at higher force levels are more stretched ( because they do not contain loops ) for pentagons . since most loops persist down to zero force levels ,
the @xmath63 for pentagons is significantly lower than for disks .
in the present paper , we discuss and describe properties of force networks in tapped particulate systems of disks and pentagons .
our analysis is based on persistent homology that allows to precisely measure and quantify a number of properties of these networks .
the persistence diagrams record the distribution and connectivity of the features ( components , loops ) that develop in the force landscape as the force threshold is decreased .
these diagrams can then be analyzed and compared by a number of different means , some of them described and used in the present work .
one of the considered concepts is the distance between the persistence diagrams that allows for their direct comparison .
the comparison can be carried out on the level of individual diagrams , allowing to compare between different configurations of nominally the same system , between different parts of a given system , or between completely different systems .
in addition , one can compare the distributions of the distances .
these comparisons allow us to identify , in a precise manner , the differences between persistence diagrams , and therefore force networks .
in addition to distances , we have defined and used other measures , such as birth times , showing at which force level features appear ; lifespans , showing how long the features persist as force threshold is modified ; and finally total persistence to describe essentially how ` mountainous ' the force landscape considered is .
the listed measures were computed both for components / clusters that could be in a loose sense related to force chains , and for loops that could be related to ` holes ' in between the force chains . the use of the outlined measures has allowed us to identify a number of features of force networks .
we use these measures , for example , to identify and explain the differences between the systems of disks exposed to different tapping intensities that lead to ( on average ) the same packing fraction .
in addition to identifying the differences between these systems , the implemented measures have also shown that the systems of disks , when exposed to low tapping intensity , evolve in a nontrivial manner , with the subsequent taps possibly correlated to the preceding ones .
we have shown that the oscillations in the measures built upon persistence diagrams are correlated with small oscillations in the packing fraction .
more generally , the finding is that if the system is tapped strongly and therefore the force network is rebuilt from scratch at each tap , the resulting force networks are similar ; however , under low tapping regime , the system ( and the resulting force network ) appears to be stuck in a certain state , and jumps out of it only infrequently .
this nontrivial finding and its consequences will be explored in more detail in our future works .
another comparison that we carried out involves tapped systems of disks and pentagons .
one important finding here is that the differences between disks and pentagons are significant when the structure of loops is considered : presence of loops is much more common for the systems of disks than for pentagons , independently of whether normal or tangential forces are considered . on the other hand ,
the differences between the persistence diagrams based on components / clusters are minor and relatively difficult to identify .
therefore , the force networks that form in tapped systems of disks and pentagons are similar when only components are considered , but significantly different when loops are included . on a more general note , it should be emphasized that the measures described here allow to directly compare force networks , and quantify the differences .
for example , we can now quantify in precise terms variability of force networks between one system and another .
this ability opens the door for developing more elaborate comparisons , measures , and also connections between the force network properties and mechanical response on the macroscale .
furthermore , the analysis that we presented here can be easily applied to the three dimensional systems , where any direct visualization may be difficult .
our future research will proceed in this direction . | in the companion paper @xcite , we use classical measures based on force probability density functions ( pdfs ) , as well as betti numbers ( quantifying the number of components , related to force chains , and loops ) , to describe the force networks in tapped systems of disks and pentagons . in the present work ,
we focus on the use of persistence analysis , that allows to describe these networks in much more detail .
this approach allows not only to describe , but also to quantify the differences between the force networks in different realizations of a system , in different parts of the considered domain , or in different systems .
we show that persistence analysis clearly distinguishes the systems that are very difficult or impossible to differentiate using other means .
one important finding is that the differences in force networks between disks and pentagons are most apparent when loops are considered : the quantities describing properties of the loops may differ significantly even if other measures ( properties of components , betti numbers , or force pdfs ) do not distinguish clearly the investigated systems . |
ngc 2146 is a peculiar spiral galaxy as seen from both the optical image and the hi line profile . although it is classified as sab by de vaucouleurs et al .
( @xcite ) , it shows a broad range of peculiarities .
measured by its far infrared flux , it is one of the 12 brightest galaxies in the sky and lies at a distance of 12.2 mpc ( @xmath0=75 km s@xmath1 mpc@xmath1 , and 1@xmath2 corresponds to 3.5 kpc ) . in optical images
there are two well - defined arms which mark the principal plane of rotation . superimposed across part of the nucleus
is an absorption band having the form of a hand , with four talon - like fingers " ( pease @xcite ) , which de vaucouleurs ( @xcite ) interpreted as being a third arm inclined to the plane of rotation .
further optical studies by benvenuti et al .
( @xcite ) , however , suggest that a simple spiral model is not adequate . in 1976 , fisher & tully mapped the region around this galaxy in the 21 cm using the nrao 91-m telescope with a resolution of @xmath3 in the n - s and e - w directions respectively and discovered an extensive `` envelope '' of neutral hydrogen around it which extends up to six holmberg radii ( 100 kpc ) .
they suggest that the abnormalities seen both optically and in the neutral hydrogen profile of the main disk are probably related to the large hi extensions observed , and believe that they might be the result of a ) tidal interaction , b ) explosion / ejection , or c ) galaxy formation .
they rule out any form of interaction between ngc 2146 and ngc 2146a , an sc typed galaxy with no evident optical abnormalities , which lies about 20 arcseconds away and is redshifted 595 km s@xmath1 with respect to ngc 2146 . however , their observations mainly aimed to trace the total extent of the hi cloud , and their resolution was not good enough to allow them to draw more definite conclusions on the causes of these abnormalities . also , the appearance of the hi envelope as a large gaseous halo around the main galaxy left room for speculations as to how large galactic halos really are , and what the impact of this might be on the qso absorption line system statistics .
ngc 2146 contains a strong radio source , @xmath4 3 kpc in size , within its nuclear region , identified with 4c 78.06 ( caswell & wills @xcite ) .
kronberg & biermann ( @xcite ) used the nrao interferometer and the vla to map the radio structure of the nuclear region .
they found that the radio center lies in the optically obscured dust lane , but it shows no evidence of a double nucleus .
the radio continuum map agrees very well with the co intensity map , and unlike the optical image , it shows a remarkable degree of symmetry ( kronberg & biermann @xcite ; jackson & ho @xcite ) .
the velocity curves measured in various lines in the optical and infrared are quite regular after allowance for extinction effects due to the dust lane ( prada et al .
kronberg & biermann ( @xcite ) suggested that a strong burst of star formation is responsible for the strong radio and infrared emission .
a @xmath5co , @xmath6co and cs study undertaken by xie et al .
( @xcite ) places the average temperature in the nuclear region about 55 k , and the average density at 2 @xmath7 10@xmath8 @xmath9 .
further evidence of a burst of start formation comes from x - ray observations carried out by armus et al .
( @xcite ) and della ceca et al .
( @xcite ) which reveal a galactic - scale outflow of gas driven by an intensive star bursting activity , referred to as a starburst - driven superwind . furthermore
, this activity can produce long - lived bending instabilities as suggested by n - body simulations carried out by griv & chiueh ( @xcite ) to explain the snake - shaped radio structure observed by zhao et al .
( @xcite ) at an angular resolution of 2@xmath10 using the vla .
observations probing the molecular content in co and h@xmath11 as well as the ionized gas content ( hii regions ) were made by young et al .
( @xcite ) .
they found a very large concentration of gas in the nucleus , confirming condon et al .
( @xcite ) earlier conclusion that this galaxy has a high star - formation rate and then , derived a mass - to - light ratio characteristic of very young stellar systems . all the above led these authors to suggest that ngc 2146 has recently undergone a collision with some other galaxy .
the existance of an extended arc of hii regions encircling the central bright region , which exhibit velocities which are 130@xmath12200 km s@xmath1 higher than those expected if they are rotating in the plane of the galaxy ( young et al .
@xcite ) , might also be interpreted as evidence of a collision . after undertaking an optical and infrared study of the galaxy , hutchings et al .
( @xcite ) found no sign of an active nucleus but did find many signs of a significant population of hot young stars in the central regions of the system .
they concluded that ngc 2146 is a merging system , now in its final stages .
the dominant galaxy is seen close to edge on , and the small companion has been stripped , leaving no sign of its nucleus .
they also note that such a scenario can also be supported by numerical simulations ( barnes @xcite ) . in order to better understand the nature of this system and decide amongst the various scenaria which have been suggested , we obtained higher resolution 21 cm maps of the hi distribution around ngc 2146 using the vla , and combined these with 21 cm observations of the nrao 91-m telescope to recover the emission on large angular scales , which the interferometric observations alone are incapable of sensing.the nature of large gaseous halos is important in the interpretation of qso absorption line spectra ( c.f rao & briggs @xcite ) , where the large cross sections implied by the fisher & tully observations ( @xcite ) would cause this kind of galaxies to intervene frequently by chance if they are common in the galaxy population .
the observations presented here consist of two parts . interferometry observations , carried out with the very large array , and single dish observations , carried out with the 300 ft nrao antenna before it collapsed . on 25,26 and 29 october 1984
, the vla was used in its most compact configuration to observe two fields , centered at @xmath13 and at @xmath14 respectively .
the center fields were chosen in such a way that their combination produces a nearly uniform response over the area between and including the two pointing coordinates ; this covers the main body of the galaxy and substantial fraction of the extended hi emission .
a total of 5.5 hours per field with 25 antennas with baselines ranging from 36 m to a little more than 1 km were used .
we observed only the right polarization in a band centered at a heliocentric velocity of 915 km s@xmath1 .
after on - line hanning smoothing , we recorded 31 independent channels ranging from 583 to 1226 km s@xmath1 with a fwhm velocity resolution of 20.75 km s@xmath1 . a broader bandwidth
`` continuum channel 0 '' to accompany the line data was calculated on - line from the average of 48 such channels centered at 915 km s@xmath1 .
the vla observational parameters are summarized in table [ obspar ] . for the flux calibration we used 3c286 , adopting a flux density of 14.88 jy according to the latest vla measurments .
the phase calibration source was 0836 + 710 for which a flux density of 4.12 jy at 1.416 mhz was determined .
both calibrators were used in extracting a measure of the shape of the passband which we applied as a correction to the observations of ngc 2146 .
[ een ] l l instrument&very large array ( vla ) + object&ngc 2146 + total observing time & 11 hours + pointing center & + south field r.a .
( 1950.0 ) & 6@xmath1512@xmath1600@xmath17 + dec .
( 1950.0 ) & + 78@xmath1810@xmath1900@xmath10 + north field r.a .
( 1950.0 ) & 6@xmath1510@xmath1600@xmath17 + dec .
( 1950.0 ) & + 78@xmath1835@xmath1900@xmath10 + flux density calibrator & 3c286 + phase calibrator & 0836 + 710 + central velocity of spectral band & 915 km s@xmath1 + configuration & d array + observed baselines ( min - max ) & 40 - 1030 m + number of antennas & 25 + number of channels & 31 + channel spacing & 20.75 km s@xmath1 + fwhm velocity resolution@xmath20 & 20.75 km s@xmath1 + total line bandwidth & 3 mhz + & 643.5 km s@xmath1 + total continuum bandwidth & 4.69 mhz + type of weighting of _ u - v _
data&natural + r.m.s .
noise per channel & 1 - 1.25 mjy + synthesized beam ( n - s e - w ) & 49@xmath21 77@xmath10 + notes + @xmath22 . on - line hanning taper mode . to optimize the sensitivity to weak extended hi emission ,
the fourier transformation was carried out using a natural weighting for the data , i.e. each visibility cell in the @xmath23 plane has been weighted according to the integration time spent on it .
this gives the best signal - to - noise ratio for detecting weak sources .
since @xmath23 tracks tend to spend more time per unit area near the @xmath23 origin , natural weighting emphasizes the data from short spacings , and produces a broad synthesized beam with an extended , low - level sideloabe .
the natural - weighted maps provide sensitive information on the larger scale and are suitable for analyzing the structure of the extended hi cloud of ngc 2146 .
however , the poor synthesized beam shape means that care must be taken in the decovolution phase of the data processing .
incorporation of the single dish spectral is required for reliable mapping of the extended structure .
the construction of the continuum subtracted channel maps for each of the two fields is a crucial step in the reduction . due to restrictions in the number of spectral channels that could be recorded in 1986 by the vla
, the correlator configuration was chosen so that it had hi emission in 29 of the 31 velocity channels . the on - line `` channel 0 '' ,
being a mean of 48 channels , 29 of which are corrupted by the strong hi emission does not represent a true continuum channel .
a true continuum map was obtained according to the following scheme , which is a weighted difference between spectral channel maps with gas and the `` channel 0 '' map : @xmath24 in this way , the continuum map was formed from the frequency band outside the channels containing hi emission .
the continuum was then subtracted from the dirty line channel maps to obtain the narrow band hi emission .
this yielded a continuum field of total flux 1.83 jy with the brightest source being the galaxy ngc 2146 ( 0.75 jy ) . during september 1987
the 300 ft nrao telescope was used to observe the area of ngc 2146 . on each day
a single drift scan was performed at a fixed declination , and spectra were recorded in short integrations . at the end of the observing session , a critically sampled map of the region surrounding ngc 2146 for merging with the interferometer data was constructed by gridding the data onto the same spatial and velocity grids as used for the vla data cubes . for the joint deconvolution performed simultaneously on the vla and 300 ft data , a gaussian beam of hpbw 11.3 arcmin in the direction n - s and 10.2 arcmin in the direction e - w was adopted for the 300 ft ( fisher & tully @xcite ) .
the continuum - free vla maps were then simultaneously cleaned , mosaiced and combined together with the single dish maps with the aips maximum entropy based task utess to form single maps for each channel . at this stage , tests were performed using other alternative approaches of deconvolution like clean algorithms alone , or combinations of clean and maximum entropy based algorithms ( implemented by vtess in aips ) , yielding similar results in terms of map quality , but making apparent the fact that the single dish data were adding to the final maps substantial flux that the interferometry maps alone were missing , due to the lack of short spacing @xmath23 coverage .
the final natural - weighted channel maps have an rms noise per channel of 1.0 to 1.25 mjy / beam , which is approximately the theoretical value ( 1.0 mjy / beam ) .
the final synthesized beam is 49@xmath2577@xmath10 in the north - south and east - west directions respectively .
the channel maps of the hi flux of the area around the optical galaxy are shown in figs .
[ fig1 ] and [ fig2 ] .
the channels containing emission at large angular distance from the galaxy , together with a map of the radio continuum at 1420mhz , appear in figs .
[ fig3 ] and [ fig4 ] .
these maps reveal elongated streams of neutral hydrogen towards the north and the south of the main galaxy , extending out up to 6 holmberg radii .
this extensive distribution of the hi has a `` tail '' morphology suggestive of a tidal interaction of ngc 2146 and a companion , which has not been identified in optical images .
the southern stream is detected at higher signal to noise ratio than the northern stream .
this occurs because the southern gas appears in only a few velocity channels , where the spatial structure is dominated by a single long arc that is nearly coincident in the three main contributing channels .
there are extensions to this arcing stream , which , instead of following the general outflow of the hi , lie in smaller arcs of gas that appear to turn eastward and separate from the stream , falling back toward the central potential of the system . further out from the tip of the southern stream
appear some small hi clouds , which do not seem to be gravitationally bound to the whole system and may be escaping towards the south east .
the stream at the north is less prominent , due in part to being more heavily resolved , both spatially and in velocity , than the southern stream .
unlike the nearly constant velocity measured in the south , there is a prominent velocity gradient across the nothern extended emission from @xmath4975 km s@xmath1 at position angle @xmath1215@xmath26 to @xmath4850 km s@xmath1 at position angle @xmath430@xmath26 .
thus , the northern stream is symmetrically placed with respect to the southern one , but is not its mirror image . instead of being concentrated in only three of our channels ( @xmath460 km s@xmath1 )
, the northern stream appears to be spread in a fan shaped outflow with line of sight velocity components ranging from @xmath460 km / s redshifted to @xmath465 km / s blueshifted as a function of position angle on the sky .
the observed hi distribution is consistent with that found by fisher & tully ( @xcite ) , given the 10@xmath2 beam ( fwhm ) of the 91 m nrao telescope .
one significant difference is that the vla detects absorption against the position of the nuclear continuum emission of the galaxy .
this absorption feature persisted throughout all our reduction , and was consistently present in both the vla maps of the north and the south part of ngc 2146 , although the position of the galaxy was at a different point in the primary beam for each case .
also , the fact that no deep bowls were seen around the hi distribution suggests that the absorption is real and not an artifact of the interferometer data reduction .
the spectrum of the absorption is shown in fig .
[ fig5 ] with 5-@xmath27 errorbars plotted for each spectral channel .
this absorption feature would certainly be missed by a single dish telescope because of its large beam , in which the absorption is diluted and overwhelmed by the integrated emission .
[ fig5 ] also shows the spectrum of ngc 2146 obtained by fisher & tully ( @xcite ) by pointing the 91 m telescope directly at the center of ngc 2146 ( dot - dashed line ) , the spectrum of the whole area containing the hi cloud made by adding up all the flux seen by the nrao 91-m telescope ( solid line ) , and the spectrum of the total hi obtained after combining the vla data with those of the nrao 91-m telescope ( solid line with xs ) .
apparently , we are still missing about 20% of the total hi flux ; maps produced from the vla data alone were missing about 50% before being combined with the single dish data .
this deconvolution procedure was successful in bringing up some fine details in the faint extended emission , especially in the northern stream of gas , where the signal to noise ratio is low due to being heavily resolved , and also in the low level emission to the east of the southern arm .
the absorption is seen against the radio center of the galaxy , which lies in the region of the optically obscuring dust , and is unresolved by our observations .
it has a velocity width of @xmath4 350 km s@xmath1 and an average optical depth of 0.03 . assuming a spin temperature of 50 k
, the measured optical depth requires an hi column density of @xmath410@xmath28 atoms @xmath29 .
the absorption is centered at the galaxy s systemic velocity and seems to be due to hi clouds rotating together with the rest of the hi seen in emission , which happen to lie in front of the strong radio continuum source of ngc 2146 , 4c 78.06 . however , the large velocity gradient seen accross its 1@xmath2 span , compared to the total width of the hi profile of ngc 2146 ( @xmath4 500 km s@xmath1 ) implies that the hi absorbing layer is not very far from the nucleus of ngc 2146 . if @xmath30 is the total hi profile velocity width of ngc 2146 ( @xmath4 500 km s@xmath1 ) , and @xmath31 is the velocity width of the absorbing layer seen against the ngc 2146 continuum emission ( @xmath4 350 km s@xmath1 ) then , assuming a circularly rotating system @xmath32 where @xmath33 is shown in fig .
[ fig6 ] . from fig .
[ fig6 ] we also see that @xmath34 where @xmath35 is the length of the absorbing layer projected on the plane of the sky ( @xmath4 1 arcsec , or 3.5 kpc ; this is an upper limit imposed by the spatial resolution of our observations ) , and @xmath36 is its distance from the nucleus . combining these two equations we estimate an upper limit for the distance of the absorbing layer from the continuum background to be @xmath37 the integrated hi flux density map is shown in figs . [ fig7 ] ( large scale ) and [ fig8 ] ( small scale ) .
we estimate the total mass for the cloud to be about @xmath38 ( @xmath0=75 km s@xmath1 mpc@xmath1 ) of which @xmath39 come from the bright region around the galaxy itself , and @xmath40 from the extended component .
furthermore , the extended distribution is not symmetric with respect to the amount of gas in the south ( @xmath41 ) and the north ( @xmath42 ) . for the above calculations a velocity component of 223 km s@xmath1 of our sun towards ngc 2146 s direction has been assumed , in order to refer the redshift velocity to the center of mass of the local group . figs .
[ fig9 ] and [ fig10 ] show the intensity weighted mean radial velocity field on both large and small scale .
points with negative flux due to absorption have been excluded from this calculation , and only emission brighter than 1.3 mjy has been taken into account .
it is clear from the velocity field that the main galaxy has a differentially rotating disk with characteristic rotational speed of about 250 km s@xmath1 .
the extended hi disk , however , appears severly lopsided as can immediately be seen from fig .
[ fig10 ] since the galaxy is seen almost edge on .
the hi disk of the main galaxy has a well defined plane of rotation , with an inclination nearly perpendicular to the sky plane .
the extended hi stream to the south , however , has very little velocity gradient along its entire length , with an average velocity that is coincident with the systemic velocity of the central galaxy .
this situation occurs when the outlying stream and the galaxy centroid lie in the plane of the sky and implies that we observe the southern stream from a preferred viewing angle , in which it appears laid out in the probable plane of the interaction , which must have been very nearly perpendicular to the disk plane of ngc 2146 . the northern stream , on the other hand ,
does show a significant velocity gradient across it , and the orientation of the stream can not be inferred in a model independent way . in fig .
[ optfig ] the total integrated hi flux density of ngc 2146 is plotted superimposed on an optical image of the galaxy from the digitized sky survey ( dss1 ) .
the morphology of the hi cloud does not show any preferential alignment with respect to the direction of ngc 2146a , and therefore shows no evidence for interaction between ngc 2146 and ngc 2146a .
the faint arm extending southwards from the top nw corner of the optical image is in the general direction of the hi southern stream .
however , the velocity gradient of hii regions of @xmath4100 km sec@xmath1 observed by young et al .
( @xcite ) across this feature is incompatible with the absence of a velocity gradient in the southern hi stream and its alignment with the plane of the sky .
it seems thus that this arm is part of the main galaxy which was disturbed from an interaction but still remained under the strong gravitational influence of ngc 2146 .
the loop structure at the se of the optical image of the galaxy shows higher velocities than those expected if it was rotating in the plane of the galaxy ( young et al . ,
@xcite ) presenting further evidence for a tidal interaction .
however , there is no apparent correspondence with any hi feature .
the complex appearance of this system may represent the aftermath of an encounter between ngc 2146 and a slightly less massive but gas rich galaxy , probably a late type lsb spiral , with a slowly rising rotation curve indicating little or no bulge ( de blok et al .
@xcite ) . numerical simulations by wallin & stuart ( @xcite ) indicate that the outcome of an interaction depends crucially on the orientation of each galaxy s rotational angular momentum vector relative to the plane of the interaction .
if the rotational angular momentum vector lies in the plane of the interaction , the galaxy comes through with very little loss of mass . on the other hand ,
if the rotational angular momentum aligns with the orbital angular momentum , a large fraction of the galaxy is stripped away .
ngc 2146 may have interacted with a gas rich `` intruder '' whose rotational angular momentum vector was oriented perpendicular to that of ngc 2146 and perpendicular to the plane of the interaction , which in this case is close to being coincident with the plane of the sky .
[ fig11 ] is a schematic diagram of the encounter . in this configuration ,
ngc 2146 has a rotational angular momentum vector lying in the interaction plane and is able to preserve its identity as a rotating disk galaxy , while the same encounter essentially destroys the intruder .
the numerical simulations of wallin & stuart ( @xcite ) predict in such cases that the mass fraction transfered from the intruder to the main galaxy can be as large as 0.5 , and the mass fraction lost by the intruder that eventually escapes from the whole system can reach up to 0.2 .
wallin & stuart used a model with restricted 3body interactions ( two large point masses binding massless test particles ) , and more realistic models with distributed mass are likely to suffer even greater destruction for the case of aligned orbital and rotational anglular momentum , as is indicated by some test cases run for us by j. gerritsen ( private communication ) using his tree - code implementation ( gerritsen @xcite ) .
the outcome of the tidal interactions is the development of gas streams on opposite sides of both galaxies .
the velocity gradient of the gas in the northern stream is consistent with the sense of rotation of ngc 2146 , which may indicate that this gas originated from the hi disk of ngc 2146 .
the gas at the southern stream , however , appears only around the systemic velocity of the system , and is likely to be the tidally dispersed remnant of the intruder .
as the galaxies swirled around each other , hi clouds released from the intruder s side opposite to ngc 2146 were boosted to escape velocity by the combination of the intruder s translational and rotational velocities . in this scenario , the outermost clouds of the intruder experience the gravitational slingshot effect familiar in interplanetary spacecraft trajectories to reach the outer solar system .
we next test the plausibility of such an encounter by exploring the timing arguments for the formation and lifetime of the southern plume near ngc 2146 . for our purposes ,
we approximate the relative speeds of the masses using the vis viva equation to describe the relative speed @xmath43 of the centers of mass for the two masses , @xmath44 and @xmath45 , as a function of their separation @xmath36 and @xmath22 , the semimajor axis of a possible elliptical orbit of the intruder about ngc 2146 .
then , @xmath46 \nonumber\\ & = & gm_{ngc2146}\left(1+\frac{m_{intr}}{m_{n2146}}\right)\left[\frac{2}{r}\right ] \nonumber\\ & \approx & ( 350 { \rm ~km~s}^{-1})^2 \left(1+\frac{m_{intr}}{m_{ngc2146}}\right)\left[\frac{15 { \rm kpc}}{r}\right ] \nonumber \end{aligned}\ ] ] where we have made the assumption that the intruder falls in from large distance so that @xmath47 and that the mass of ngc 2146 can be approximated by adopting the rotational velocities of @xmath48250 km s@xmath1 that are measured in the outskirts of ngc 2146 at @xmath4915 kpc to obtain @xmath50 . for purposes of this illustration
, we consider an hi - rich , `` lsb '' intruder , with @xmath51 and @xmath52 , having a gradually rising rotation curve that reaches @xmath53 km s@xmath1 at the outer edge of the galaxy s disk . under these assumptions , the relative speeds of the centers of mass for
the two galaxies will be @xmath54 km s@xmath1 for a grazing encounter where @xmath55 kpc .
four - fifths of the encounter speed ( @xmath56222 km s@xmath1 ) is carried by the center of mass of this intruder at a distance of 24 kpc from the system barycenter . by the time of closest approach , the intruder will have already become tidally distorted , and as it passes pericenter , simple tidal force arguments show that it will be unbound throughout its disk , provided it has a slowly rising rotation curve . however , in order to estimate the `` launch speed '' of debris in the outskirts of the intruder ( trajectory a in figure 12 ) , we consider the effect of adding the @xmath57 of the intruder to the speed of its center of mass . under these assumptions ,
gas clouds at the far side of the intruder from ngc 2146 are traveling at @xmath4346 km s@xmath1 at a distance of 39 kpc from the barycenter .
we estimate the escape velocity from this distance by lumping @xmath58 at the barycenter to find @xmath59 km s@xmath1 , which is roughly 100 km s@xmath1 less than the speed of the outer gas clouds .
once the process of dismantling the intruder begins , the mass lost lowers the binding energy , allowing the intruder to be shredded throughout .
gas clouds on the inside of the intruder closest to the system barycenter , of course , find that whatever rotational component of velocity they still carry is counter to the orbit around ngc 2146 .
the dynamics in this region are clearly complicated , so that subtracting the 125 km s@xmath1 rotational speed from the 222 km s@xmath1 intruder center of mass speed to leave @xmath4100 km s@xmath1 forms only a rough guess . however , the implication is still that the inner gas is progressing in the same direction as the bulk of the material from the dismemberment of the intruder . as a result
, we picture the southern hi plume to be the result of tidal shredding of a high angular momentum lsb galaxy .
the test simulation by gerritsen ( private communication ) confirms that this could be accomplished without enhancing star formation in the intruder . in order to make some statement about the time span since closest approach
, we make the following simplifying assumptions : first , the high speed clouds from the outskirts of the intruder escape ( trajectory a in figure 12 ) from the system by overcoming the gravitation potential created by the the sum of the two original masses , concentrated at the system barycenter .
second , the remnants such as the hi spur running nearly east - west at declination @xmath60 find their acceleration dominated by the mass of ngc 2146 alone .
if we assume that the gas in this spur is located at the turn - around point of its trajectory , we can estimate the major axis of its orbit and deduce the time span required for it to return to closest approach to ngc 2146 .
the spur falls @xmath61 or 49 kpc from the center of ngc 2146 . assuming these clouds were launched from a distance of @xmath62 kpc on the opposite side of ngc 2146 , this trajectory ( b in figure 12 ) can be approximated by an ellipse of major axis @xmath63 kpc . using the rotational velocity @xmath64250 km s@xmath1 for ngc 2146 at @xmath4915 kpc
, we deduce that the orbital period @xmath65 would be @xmath66 as a function of @xmath22 .
thus , half of the period for @xmath67 kpc is @xmath40.8 gyr .
this forms an estimate of the time elapsed since the closest approach .
the gas clouds ejected along trajectory a have reached @xmath68 or 180 kpc during 0.8 gyr , for an average speed of 225 km s@xmath1 .
if the clouds were launched with the 346 km s@xmath1 estimated above , then they will have decelerated along their path , consistent with this estimate for average speed . in this scenario ,
the central regions of the destroyed intruder are likely to lie somewhere along the south stream .
indeed , there is a concentration of gas , located at declination 78@xmath2604@xmath2 , where the south stream has a kink ( most clearly seen in channels with velocities of 853 and 832 km s@xmath1 in fig .
. the concentration has galaxy sized dimensions and a velocity gradient of 60 km s@xmath1 across its 30 kpc extent , contains @xmath4 1.5@xmath710@xmath69 in neutral hydrogen , and has an hi column density of @xmath4 2@xmath710@xmath70 atoms @xmath29 , which is high enough to trigger star formation that might be detectable in deep ccd images .
all these make this object a very good candidate of being a remnant of the long sought companion of ngc 2146 .
clearly more observations need to be undertaken to clarify the nature of this object , and test the validity of this scenario .
another plausible scenario for this system is that it is at the final stage of merging and the small companion has been completely stripped off of its gas leaving no sign of its nucleus , as suggested by hutchings et al .
( @xcite ) and lisenfeld et al .
( @xcite ) . in view of the support that numerical simulations can provide to such mergers ( barnes @xcite ) and
if the above mentioned hi concentration at the south does not provide evidence for a companion nucleus around ngc 2146 this seems also a very attractive possibility .
the ngc 2146 system provides another example to add to those of hibbard & van gorkom ( @xcite ) where galaxy - galaxy interactions inject galactic gas into galactic halo regions , as well as ejecting gas with sufficient velocity that it can escape to the intergalactic medium .
depending on the mass ratio of the interacting galaxies , the relative inclinations of their disks and the impact parameter of the encounter the mass lost in the interaction can be up to 60% of the mass of the companion galaxy ( wallin & stuart @xcite ) .
this gas contributes to the enrichment of the intergalactic medium in metals and if such events had been very common at some earlier epochs they may help towards finding the objects responsible for the metal absorption - line systems seen in abundance in qso spectra . which of the above scenaria is correct remains yet undetermined .
numerical simulations based on the morphological and kinematical information presented may give us a clearer picture of how the system looked a billion years ago .
nevertheless it is striking that the final result of this collision is a morphologically classified spiral galaxy , although both the galaxy ngc 2146 and its surrounding hi cloud seem to have contained approximately equal amounts of neutral hydrogen .
this system does provide evidence that mergers of two significant systems can be important events in the history of spiral as well as elliptical galaxies .
the encounter will affect the host galaxy for a long time into the future , since the outlying gas residue is likely to fall into the galaxy both from north and south meeting the rotational plane of the galaxy at a a significant angle of inclination to its internal orbital plane ; this may provide a trickle of missaligned angular momentum for a long time as gas on different trajectories turns around and falls back . in this sense ,
the effect of this collisional debris will resemble binney s `` cosmic drizzle '' ( @xcite ) as a mechanism for creating long lived warps in large , isolated galaxies .
by combining single dish observations and d - configuration vla data we have produced sensitive , high quality images of the starburst galaxy ngc 2146 .
these images reveal elongated streams of neutral hydrogen extending out up to 6 holmberg radii towards both the north and the south of the main galaxy .
the hi morphological and kinematical picture suggests that this galaxy underwent a strong interaction with a lsb companion , which was destroyed during the encounter .
although quite disturbed , ngc 2146 retained the morphological characteristics of a spiral galaxy .
analysis of the trajectory of the outlying gas suggests that the interaction was in its most violent phase about 0.8@xmath710@xmath71 years ago .
the infall of the outlying gas to ngc 2146 will continue for a similar timespan preserving a long lived warp in the disk of this galaxy .
we wish to acknowledge the national astronomy and ionosphere center at the arecibo observatory for their hospitality during part of the data reduction .
the national radio astronomy observatory ( nrao ) is operated by associated universities , inc . under cooperative agreement with the national science foundation .
this research has made use of the nasa / ipac extragalactic database ( ned ) which is operated by the jet propulsion laboratory , caltech , under contract with the national aeronautics ans space administration .
this work has been supported by national science foundation grant ast 91 - 19930 .
armus l. , heckman t. m. , weaver k. a. , lehnert m. d. , 1995 , apj 445 , 666 barnes j. , 1990 , iau colloquium no 134 benvenuti p. , capaccioli m. , odorico s. , 1975 , a&a 41 , 91 binney j. , 1990 , in : dynamics and interactions of galaxies , eds wielen r. , springer verlag , berlin , p. 328 de blok w. j. g. , mcgaugh s. s. , van der hulst j. m. , 1996 , mnras 283 , 18d caswell j. l. , wills d. , 1967 , mnras 135 , 231 condon j. j. , condon m. a. , gisler g. , puschell , j. j. , 1982 , apj 252 , 102 della ceca r. , griffiths r. e. , heckman t. m. , lehnert m. d. , weaver k. a. , 1999 , apj 514 , 772 fisher j. r. , tully r. b. , 1975 , a&a 44 , 151 fisher j. r. , tully r. b. , 1976 , a&a 53 , 397 gerritsen j. p. e. , 1997 , ph.d .
thesis , university of groningen griv e. , chiueh t. , 1998 , apj 503 , 186 hibbard j. e. , van gorkom j. h. , 1996 , aj 111 , 655 hutchings j. b. , neff s. g. , stanford s. a. , unger s. w. , 1990 , aj 100 , 60 jackson j. , ho p. t. p. , 1988
, apj 324 , l5 kronberg p. p. , biermann p. , 1981 ,
apj 243 , 89 lisenfeld u. , alexander p. , pooley g.g . , wilding t. , 1996 , mnras 281 , 301 pease f. g. , 1920 , apj 51 , 276 prada f. , beckman j. e. , mc keith c. d. , castles j. , greve a. , 1994 , apj 423 , l35 rao s. , briggs f. h. , 1993 , apj 419 , 515 de vaucouleurs g. , 1950 , ann
. astrophys .
13 , 362 de vaucouleurs g. , de vaucouleurs a. , corwin h. g. , 1976 , second reference catalog of bright galaxies , university of texas press , austin wallin j. f. , stuart b. v. , 1992 , apj 399 , 29 xie s. , young j. , schloerb f. p. , 1994
, apj 421 , 434 young j. s. , claussen m. j. , kleinmann s. g. , rubin v.c . , scoville n. , 1988 , apj 331 , l81 zhao j. h. , anantharamaiah k. r. , goss w. m. , viallefond f. 1996 , apj 472 , 54 | ngc 2146 is a peculiar spiral galaxy which is currently undergoing a major burst of star formation and is immersed in a extended hi structure that has morphological and kinematical resemblence to a strong tidal interaction .
this paper reports aperture synthesis observations carried out in the 21 cm line with the very large array ( vla ) of two fields positioned to optimally cover the hi streams to the north and south of the galaxy , along with a 300 ft total power spectral mapping program to recover the low surface brightness extended emission .
the observations reveal elongated streams of neutral hydrogen towards both the north and the south of the optical galaxy extending out up to 6 holmberg radii .
the streams are not in the principle plane of rotation of the galaxy , but instead are suggestive of a tidal interaction between ngc 2146 and a lsb companion that was destroyed by the encounter and remains undetected at optical wavelengths .
part of the southern stream is turning back to fall into the main galaxy , where it will create a long - lived warp in the hi disk of ngc 2146 .
analysis of the trajectory of the outlying gas suggests that the closest encounter took place about 0.8 billion years ago and that infall of debris will continue for a similar time span . |
one of the central problems in applied mathematics is the analysis of signals .
usually signals are modelled by functions in suitable functions spaces ( e.g. , @xmath4 or sobolev spaces ) and they might be given explicitly or implicitly as the solution of an operator equation .
in most applications , the signal is transformed via a mapping into a suitable parameter space where it is easier to extract the information of interest . by discretization ,
one obtains suitable building blocks that give rise to a discrete representation of the signal and can be used to decompose , compress and process the signal . over the years , many different transforms have been derived in response to particular problems , including the wavelet and gabor transforms .
representation theory , however , gives a general approach to construct continuous transforms for @xmath4-functions , and coorbit space theory allows both to extend these transforms to more general function spaces and to provide discrete systems .
indeed , it was shown that virtually all well - known transforms used in signal analysis can be derived from this general setting . in this sense ,
coorbit space theory serves as a common thread in the jungle of all possible signal transformations .
nevertheless , as will be explained below , the classical coorbit space setting relies on specific assumptions that might be hard to verify in practice .
the purpose of this paper is to investigate how to weaken these basic assumptions with the goal of extending the applicability of this framework to a much larger class of problems .
coorbit space theory was originally introduced by h. feichtinger and k. grchenig in a series of papers in 1988 - 89 @xcite . by means of this theory , given a square integrable representation ,
it is possible to construct in an efficient and systematic way a full scale of smoothness spaces where the smoothness of a function is measured by the decay of the so - called _
voice transform_. for any unitary representation @xmath5 of a locally compact topological group @xmath0 on a hilbert space @xmath6 and a fixed @xmath7 , the voice transform @xmath8 is the map assigning to @xmath9 the corresponding transform @xmath10 as @xmath11 ranges in @xmath0 .
evidently , @xmath12 is a function on the group .
since @xmath6 is often a hilbert space of functions , the voice transform connects two function spaces , that is , it maps signals to functions on the group .
coorbit space theory has been very successful in many ways and has given rise to a wealth of results , enabling to derive a very large family of smoothness spaces as coorbit spaces , including both classical functions spaces and new ones .
in particular , the classical besov spaces are derived as coorbit spaces where smoothness is measured by the decay of the wavelet transform , i.e. , the voice transform associated with the affine group .
similarly , the well - established class of modulation spaces corresponds to the family of coorbit spaces where smoothness is measured by the decay of the gabor transform , i.e. , the voice transform associated with the weyl - heisenberg group ( cf .
@xcite ) . as another example ,
let us mentions the @xmath13modulation spaces @xcite which can be interpreted as coorbit spaces related to group representations modulo quotients @xcite .
another advantage of coorbit space theory is to provide atomic decompositions and banach frames for the coorbit spaces , through a procedure which generates discrete function systems by discretization of the group representation .
this is important since it provides a way to understand the properties of discrete signal representations through the group theoretic properties of their corresponding continuous voice transforms . in recent years
, a new generation of multiscale transforms has emerged in applied harmonic analysis , such as the shearlet and curvelet transforms , which were introduced to overcome the limitations of the traditional multiscale framework in multi - dimensional setting with high efficiency @xcite .
recent results have shown that the continuous shearlet transform , in particular , stems from a square integrable group representation of the so - called _ full shearlet group _ @xcite . by applying the coorbit space theory to this setup
, it is possible to define some useful anisotropic smoothness spaces via the decay properties of the shearlet transform and to relate these spaces to other well - known function spaces @xcite .
this has stimulated the investigation of a larger class of group representations , primarily those arising from the restriction of the metaplectic representation to a class of triangular subgroups of the symplectic group @xcite .
this class includes many known cases of interest in signal analysis and gives rise to several new examples , such as the schrdingerlets that we discuss in this paper .
yet another potential extension of this framework is the general context of the so - called mock metaplectic representations , introduced in @xcite .
however , the classical coorbit space theory appears to be too restrictive to deal with this more general class of group representations and the corresponding voice transforms and function spaces .
let us recall that the classical coorbit space theory _ la _ feichtinger - grchenig makes the following two assumptions : + ( fg1 ) the _ kernel _ @xmath14 , that is , the voice transform of the admissible vector itself , is an absolutely integrable function on the group .
+ ( fg2 ) the representation is assumed to be irreducible .
+ a major part of this paper is concerned with replacing ( fg1 ) by some weaker condition .
the fundamental concepts will be presented in sections 2 and 3 .
first of all , let us mention that the problem of removing the integrability condition has already been addressed by j. christensen and g. lafsson in @xcite .
classically , the reservoir of test functions is obtained by taking the functions whose voice transform is in @xmath15 , hence it is a banach space in a natural way . in the papers @xcite , the reservoir is a fixed frchet space @xmath16 densely embedded into @xmath6 .
the basic example is the set of @xmath17 vectors for the representation .
the approach that we consider in this paper is indeed similar to @xcite , but features an important new datum that we call the _ target space _ @xmath18 .
this is a frchet space of functions on @xmath0 and plays the role of @xmath15 in the classical setup . in our theory , the reservoir @xmath16 is the set of functions whose voice transform is in @xmath18 .
in section 4 , we provide a new model for the target space , namely @xmath19 for this choice we are able to produce concrete cases arising from triangular subgroups of @xmath20 , notably the case of the so - called _ schrdingerlets _ , discussed in section 4.3 . as a toy example , in section 4.2 we also consider the case of the band - limited functions with @xmath21 but not in @xmath22 . ] . as for assumption ( fg2 ) ,
most of the classical coorbit space theory can be carried out also in the reducible case . in the irreducible case ,
it is possible to show that the construction of the coorbit spaces is independent of the choice of the admissible vector .
exactly this property is lost in the reducible case , as showed by an important example in @xcite . neither in @xcite nor in our setting irreducibility is needed . in section [ av ]
we give a reasonable description of admissible vectors leading to the same coorbit spaces .
finally , in section [ l1 ] we present a detailed account of the extent to which the classical @xmath22 theory can be developed without the assumption of irreducibility .
let us briefly describe in some detail the main features of our approach .
the starting ingredients are a unitary reproducing representation @xmath5 of a locally compact second countable group @xmath0 on a separable hilbert space @xmath6 and an admissible vector @xmath7 .
next , the frchet spaces @xmath16 and @xmath1 come into play , and their roles in the theory can be described by the following very basic conceptual picture @xmath23 where @xmath24 denotes the space of measurable functions on @xmath0 .
thus , @xmath25 and @xmath18 are frchet spaces that embed into @xmath6 and @xmath24 , respectively , and should therefore be thought of as signals and functions on the group , respectively .
the space @xmath1 is a free choice , as long as one can embed it continuously into @xmath24 in such a way that some basic properties are satisfied ( assumptions [ h1 ] and [ h2 ] ) .
the space @xmath25 will serve as the reservoir of _ test functions _ and it is defined as the subset of @xmath6 consisting of those vectors whose voice transform belongs to @xmath26 .
test functions are modeled in terms of their voice transforms and the latter ones constitute the true degree of freedom in the construction , the target space @xmath18 . once the basic structures are laid out , one then follows the lines of coorbit space theory and defines first the distributions @xmath27 and then coorbits associated to banach spaces of functions .
technically speaking , our theory is determined by the data set @xmath28 , where : * @xmath0 is a locally compact second countable topological group ; * @xmath6 is a separable hilbert space ; * @xmath5 is a continuous unitary reproducing representation of @xmath0 on @xmath6 ; * @xmath7 is an admissible vector ; * @xmath1 is a frchet space continuously embedded via @xmath29 into @xmath24 ; * @xmath30 is a banach space continuously embedded in @xmath31 and left invariant .
the main ancillary objects attached to them are : * the voice transform @xmath32 ; * the reproducing kernel @xmath33 ; * the reproducing kernel space @xmath34 ; * the space of test functions @xmath35 ; * the space of distributions @xmath36 ; * the extended voice transform @xmath37 ; * the coorbit space @xmath38 .
these data are assumed to satisfy several assumptions that are made both for the target space @xmath1 and for the model space @xmath30 .
assumption [ h1 ] asks that the kernel @xmath39 is in @xmath1 , which substitutes the classical integrability condition @xmath40 , and that @xmath41 whenever @xmath42 .
this second requirement has a twin version for @xmath30 , namely assumption [ h5 ] , and it is trivially satisfied in the case @xmath43 because @xmath39 is bounded .
assumption [ h2 ] and assumption [ h6 ] ask that the product of any voice transform and any `` good '' function in @xmath44 ( or in @xmath45 ) is in @xmath15 .
assumption [ h3 ] ensures that the extended voice transform is injective .
this is a necessary condition to reconstruct a distribution from its voice transform .
finally , assumption [ h4 ] requires that the reproducing formula extends to all distributions .
in particular , we prove in proposition [ repr - formula ] that assumption [ h4 ] holds true if @xmath1 is reflexive and @xmath46 .
our theory is succesful in the sense that : it provides a workable substitute for the classical integrability condition @xmath40 ; it contains the classical coorbit space theory even for non irreducible representations ; it applies to several interesting examples ; it is consistent with the recent theory developed in @xcite .
in this section , we recall some properties of the frchet spaces that are relevant to the main objects of our theory , namely the _ target space _ @xmath18 and the space @xmath16 of _ test signals _ that will be defined in the next section . we introduce abstract spaces @xmath47 and @xmath48 .
the space @xmath47 must be interpreted as modeling a subspace of the hilbert space @xmath6 , hence of signals , but possibly with a different topology .
its properties will be used primarily for the test space @xmath16 , but also for @xmath6 itself .
similarly , the space @xmath48 should be thought of as an abstract model of a frchet space of functions on the group .
the results proved for @xmath48 will be primarily applied to the target space @xmath18 , which in many examples is a genuine frchet space but not a banach space , but will also be useful for @xmath49 , @xmath50 and , most notably , for @xmath51 , the space of functions used to define coorbit spaces . from this point of view
, our theory indicates that it is possible to develop a useful coorbit space theory assuming that @xmath30 is a frchet space rather than the more common choice of a banach space .
this further extension , however , is beyond the scope of this article , and we content ourselves with the classical case in which @xmath30 is banach space .
we now introduce the basic notation and recall some elementary properties .
further technical results are recalled in section [ notation ] . throughout this paper , @xmath0 ,
denotes a fixed locally compact second countable group with a left haar measure @xmath52 and @xmath53 is its modular function .
we write @xmath54 instead of @xmath55 and denote the classical spaces of complex functions on @xmath0 as follows :
[ cols= " < , < " , ] the space @xmath31 is a metrizable complete topological vector space with respect to the topology of convergence in measure ( see section [ notation ] ) .
the norm of @xmath56 and the scalar product between @xmath57 are denoted by @xmath58 and @xmath59 , respectively .
the space @xmath60 is a frchet space with respect to the topology defined by the family of semi - norms @xmath61 where @xmath62 runs over the compact subsets of @xmath0 ( see section [ notation ] ) .
we denote by @xmath63 and @xmath64 the left and right regular representations of @xmath0 on @xmath24 , namely @xmath65 for all @xmath66 , all @xmath67 and almost all @xmath68 . both @xmath63 and @xmath64 leave @xmath60 and each @xmath69 invariant , and @xmath63 is equicontinuous both on @xmath60 and on each @xmath69 . in section [ reps ]
we recall the main properties of the representations acting on frchet spaces . for general background on representations
the reader is referred to @xcite . for all @xmath67
we denote by @xmath70 the element in @xmath24 given by @xmath71 for almost all @xmath72 ( see section [ notation ] ) . given two functions @xmath73
, we say that the convolution @xmath74 exists if for almost all @xmath66 the function @xmath75 is in @xmath15 .
we write @xmath76 and we have that @xmath77 ( see section [ convolution ] ) . in
what follows , @xmath5 denotes a fixed strongly continuous unitary representation of @xmath0 acting on the separable hilbert space @xmath6 and @xmath78 a fixed vector in @xmath6 .
we stress that @xmath5 is not assumed to be irreducible , nor , at this stage , reproducing .
as it is customary , the _ voice transform _ associated to the these data is the map @xmath79 it intertwines @xmath5 and @xmath80 , that is @xmath81 for all @xmath66 .
the corresponding kernel is given by @xmath82 it enjoys the basic symmetry property @xmath83 for all @xmath84 the _ fourier transform _
@xmath85 is the bounded operator on @xmath6 defined by @xmath86 for all @xmath87 .
note that , with the choice @xmath88 , we get @xmath89 in this section , we consider a space @xmath90 of functions on @xmath0 and we study the properties of the convolution operator @xmath91 from @xmath90 into @xmath92 .
in particular , we introduce the subspace @xmath93 of those functions which are left fixed by the convolution operator . in the theory of reproducing representations , @xmath90 is the hilbert space @xmath94 ; in the theory developed by h. feichtinger and k. grchenig it is a weighted version of @xmath15 ; in our setting it is the target space @xmath1 .
we assume that @xmath90 is a frchet space with a continuous embedding @xmath95 . with slight abuse of notation , given @xmath96 ,
we denote by @xmath97 a @xmath52-measurable function such that for almost every @xmath66 , @xmath98 .
further , we assume that there exist a. a continuous involution @xmath99 on @xmath90 such that @xmath100 ( so that @xmath101 ) ; b. a continuous representation @xmath102 of @xmath0 acting on @xmath90 for which @xmath103 so that @xmath104 and @xmath105 .
standard examples of spaces satisfying the above assumptions are the @xmath106-spaces or their weighted versions .
other important examples are the space of @xmath17 functions , if @xmath0 is a lie group , or the space of rapidly decreasing functions , whenever this notion makes sense .
the space @xmath107 is a subspace of @xmath31 , stable under complex conjugation and @xmath63-invariant .
clearly , we could identify @xmath90 with @xmath107 avoiding the cumbersome map @xmath29 .
however , we want to stress that @xmath90 has its own topology , which is not necessarily the topology of @xmath107 , that is , the relative topology as a topological subspace of @xmath31 . in order to clarify the role of the two topologies
, we shall not identify @xmath90 with @xmath107 . since @xmath29 is continuous from @xmath90 into @xmath31 , for any sequence @xmath108 converging to an element @xmath109 in @xmath90
, there exists a subsequence @xmath110 such that @xmath111 converges to @xmath112 for almost all @xmath66 ( see for details ) .
we denote by @xmath113 the topological dual of @xmath90 . for each @xmath114 , the map @xmath115 defines a continuous anti - linear function on @xmath90 , which we denote by @xmath116 .
the map @xmath117 is a linear isomorphism of @xmath113 onto the anti - dual @xmath118 , the space of anti - linear continuous forms on @xmath90 . in what follows
, we identify @xmath118 with @xmath113 .
observe that the map @xmath119 is a sequilinear form on @xmath120 , linear in the first entry and anti - linear in the second .
the kthe dual of @xmath90 is defined by @xmath121 it is closed under complex conjugation and its elements can be regarded as anti - linear forms on @xmath90 , as shown by the next lemma . here and below , we fix a countable fundamental system @xmath122 of saturated semi - norms in @xmath90 .
[ ltre ] given @xmath123 , the map @xmath124 is a continuous anti - linear form , that is , an element of @xmath113 , which we denote again by @xmath125 .
the representation @xmath63 leaves @xmath126 invariant and for all @xmath66 @xmath127 finally , there exist a constant @xmath128 and a semi - norm @xmath129 in the fundamental saturated family @xmath130 such that for all @xmath96 @xmath131 by definition , for each @xmath96 the function @xmath132 is in @xmath15 .
we claim that the linear map @xmath133 is continuous . since both @xmath90 and @xmath15 are separable metrizable vector spaces , by the closed graph theorem ( corollary 5 of chapter i.3.3 of @xcite ) it is enough to show that the graph of @xmath134 is sequentially closed in @xmath135 .
take a sequence @xmath136 in @xmath90 converging to @xmath109 in @xmath90 and such that @xmath137 converges to @xmath138 in @xmath15 .
since both @xmath90 and @xmath15 are continuously embedded in @xmath31 , possibly passing to a subsequence , we can assume that both @xmath139 and @xmath140 converge to @xmath112 and @xmath141 , respectively , for almost every @xmath11 .
hence for almost all @xmath66 @xmath142 that is , @xmath143 in @xmath15 . hence @xmath134 is continuous , as well as the anti - linear form @xmath144 we now prove that @xmath63 leaves @xmath126 invariant .
indeed , given @xmath145 and @xmath96 @xmath146 where the last integral is finite since @xmath147 .
hence @xmath148 .
the same string of equalities gives .
finally , the last formula follows directly from the continuity of @xmath134 . in general , @xmath126 is a proper subset of @xmath113 as the following example clarifies .
take @xmath149 with @xmath150 $ ] .
then @xmath151 , where @xmath152 is the dual exponent of @xmath153 , so that @xmath154 for all @xmath155 , but of course @xmath156 .
the next proposition shows that , if the kernel @xmath39 belongs to the kthe dual of @xmath90 , then for all @xmath157 the convolution @xmath158 exists .
furthermore , we introduce the subspace @xmath159 whose elements are those reproduced by convolution with @xmath39 on the right . in the following statement
@xmath92 is endowed with the topology of the compact convergence .
[ uno ] assume that @xmath160 .
then : a. for all @xmath96 , @xmath158 exists everywhere , it is a continuous function and , for all @xmath66 , @xmath161 b. the map @xmath162 is continuous from @xmath90 to @xmath92 ; c. the set @xmath163 is an @xmath102-invariant closed subspace of @xmath90 , and therefore it is a frchet space respect to the relative topology .
notice that , in view of , @xmath164 if and only if @xmath165 .
since @xmath63 leaves @xmath126 invariant ( lemma [ ltre ] ) , for all @xmath66 we have @xmath166 and , for all @xmath96 , @xmath167 hence @xmath168 exists and the first equality of holds true .
the change of variables @xmath169 proves the second equality of .
since the involution and @xmath170 are continuous , implies that @xmath158 is a continuous function . to prove b ) , fix a compact subset @xmath171 . by , since @xmath172 , there exist two semi - norms @xmath173 in the fundamental saturated system @xmath130 and constants @xmath174 and @xmath175 such that @xmath176 where the last inequality follows from the fact that @xmath177 is equicontinuous since @xmath178 is compact ( see section [ reps ] ) . as for c ) ,
since @xmath90 is a metrizable vector space , it is sufficient to prove that @xmath93 is sequentially closed .
take a sequence @xmath136 in @xmath93 converging to @xmath96 . possibly passing to a subsequence
, we can assume that there exists a negligible set @xmath179 such that for all @xmath180 @xmath139 converges to @xmath112 .
furthermore , possibly changing @xmath179 , we can also assume that , for all @xmath181 and @xmath180 , @xmath182 .
hence , given @xmath180 , by b ) we have @xmath183 hence @xmath184 in @xmath31 , that is @xmath185 .
finally , given @xmath66 and @xmath185 , by in the appendix @xmath186 that is @xmath187 .
in what follows , for each @xmath188 , we choose the continuous everywhere defined function @xmath158 as representative of @xmath109 , so that for all @xmath66 @xmath189 we are interested in extending the voice transform @xmath8 from @xmath6 to some bigger space , namely the dual of a frchet space @xmath190 which is continuously embedded into @xmath6 , in such a way that the duality relation still holds true . in the classical coorbit space theory and in our setting
, @xmath190 is the space of test functions @xmath16 , which in @xcite is the basic object on which the theory is developed .
we fix a frchet space @xmath190 together with a continuous representation @xmath191 of @xmath0 acting on @xmath190 and a continuous embedding @xmath192 intertwining @xmath191 and @xmath5 .
as above , we identify the dual @xmath193 and the anti - dual @xmath194 .
we are interested in the transpose @xmath195 , defined as usual by @xmath196 we assume that @xmath197 and , with slight abuse of notation , we regard @xmath78 as an element both in @xmath190 and in @xmath6 . hence , we define the _ extended voice transform _ by @xmath198 which intertwines the contragredient representation @xmath199 with the left regular representation @xmath63 .
hereafter we establish some useful properties of the abstract space @xmath190 , that will be applied to the space @xmath16 of test functions in section [ testanddist ] .
a basic requirement on @xmath200 is that it must be injective . in the next lemma
some standard equivalent conditions are established .
[ b ] the following facts are equivalent : a. [ b1 ] the map @xmath200 is injective ; b. [ b2 ] the set @xmath201 is total in @xmath190 ; c. [ b3 ] the set @xmath201 is total in is just @xmath190 endowed with the weak topology . ]
@xmath202 .
define @xmath203 as the closure in @xmath190 of the linear span of @xmath201 . since the linear span of @xmath201 is convex , @xmath203 is convex ( proposition 14 chapter ii.2.6 of @xcite ) .
then @xmath203 is a closed convex subset , and so it is also weakly closed ( proposition 1 chapter iv.1.1 of @xcite ) .
therefore @xmath203 is also the closure in @xmath202 of the linear span of @xmath201 .
this proves the equivalence between [ b2 ] ) and [ b3 ] ) .
next , assume [ b2 ] ) , that is @xmath204 . if @xmath205 is such that @xmath206 , then @xmath207 for all @xmath208 , hence @xmath209
. therefore [ b2 ] ) implies [ b1 ] ) .
finally , suppose @xmath210 .
then , as a consequence of the hahn banach theorem ( @xcite , proposition 2 , p.180 ) , there exist @xmath205 and @xmath211 such that @xmath207 for all @xmath212 and @xmath213 , so that @xmath206 but @xmath214 . therefore [ b1 ] ) implies [ b2 ] ) . by definition , @xmath78 is a cyclic vector for the representation @xmath191 if condition [ b2 ] ) holds true .
notice that the cyclicity for @xmath5 does not imply the cyclicity for @xmath191 , since the topology of @xmath190 is finer than the topology of @xmath6 .
the topological dual of @xmath47 comes now into play and will be denoted by @xmath193 .
when topological properties are involved , @xmath193 is understood to have the topology of the convergence on the bounded subsets of @xmath47 .
we will write @xmath215 to stress when @xmath193 is rather thought with the topology of the simple convergence ( compare section [ scalar_integration ] ) .
the following proposition shows that , for any function @xmath67 satisfying a suitable integrability condition , it is possible to define an element in @xmath193 which plays the role of the fourier transform of @xmath109 at @xmath78 .
it is useful to compare our assumption with condition ( r3 ) in @xcite .
[ due ] take @xmath67 and assume that @xmath216 then there exists a unique @xmath217 such that , for all @xmath218 , @xmath219 for any such @xmath109 , we have @xmath220 finally , assume that @xmath221 exists , is equal to @xmath222 and that @xmath223 exists
. then @xmath224 define the map @xmath225 by @xmath226 .
since for all @xmath227 @xmath228 by we know that the map @xmath229 is scalarly @xmath52-integrable .
since @xmath190 is a frchet space , then it satisfies the ( gdf ) property
. then theorem [ gelfand - dunford ] applies , showing that the scalar integral @xmath230 exists and belongs to @xmath193 ( see section [ scalar_integration ] ) .
we set @xmath231 and , by definition of scalar integral , holds true for all @xmath227 .
also , for all @xmath66 , @xmath232 finally , under the ongoing assumptions , in the appendix implies that @xmath233 , so that is a direct consequence of .
if is satisfied , we say that the fourier transform of @xmath109 at @xmath78 exists in @xmath193 or , simply , that @xmath217 exists .
condition is actually both necessary and sufficient to define @xmath234 as an element of @xmath193 .
the next lemma ensures that the voice transform is reproduced by convolution .
[ rsei ] assume that the extended voice transform is injective and take @xmath235 .
the following assertions are equivalent : a. [ c3 ] @xmath236 exists and satisfies the reproducing formula @xmath237 b. [ e2 ] for all @xmath66 , the map @xmath238 is in @xmath15 and @xmath239 if the fourier transform of @xmath240 at @xmath78 exists in @xmath193 , _
i.e. _ the map @xmath241 is scalarly integrable , then a ) and b ) are also equivalent to each of the following assertions : a. [ c1 ] @xmath242 ; b. [ c2 ] the reconstruction formula @xmath243 holds true weakly .
the equivalence between [ c3 ] ) and [ e2 ] ) is just the definition of @xmath200 and @xmath39 .
taking into account that @xmath200 is injective , the equivalence between [ c3 ] ) and [ c1 ] ) follows from with @xmath244 .
the equivalence between [ c1 ] ) and [ c2 ] ) is just the definition of scalar integral . in the next proposition
we assume that the fourier transform @xmath234 exists in @xmath193 for all @xmath96 , where @xmath90 is a frchet space satisfying the assumptions of section [ subsec : target ] . with slight abuse of notation ,
we write @xmath234 instead of @xmath245 .
we define the coorbit space @xmath246 [ cuno ] take @xmath247 , @xmath248 and @xmath249 as above .
assume that for all @xmath96 and all @xmath227 @xmath250 which may be rephrased as @xmath251 .
then : a. the fourier transform of any @xmath252 at @xmath78 exists in @xmath193 , so does the convolution @xmath253 , and @xmath254 b. the space @xmath93 , defined by , is an @xmath102-invariant closed subspace of @xmath90 , @xmath255 is a @xmath199-invariant subspace of @xmath193 , and @xmath200 intertwines @xmath199 with @xmath63 ; c. the map @xmath256 is continuous from @xmath257 into @xmath258 ; d. if @xmath200 is injective and the reproducing formula holds for all @xmath259 , then + @xmath260 + hence , @xmath200 is a bijection of @xmath255 onto @xmath261 and therefore it induces a bijection , denoted again by @xmath200 , from @xmath255 onto @xmath93 , whose inverse is the fourier transform at @xmath78 .
item a ) is a direct consequence of proposition [ due ] .
item b ) is due to proposition [ uno ] .
the invariance property of @xmath255 is a consequence of the fact that @xmath262 for all @xmath66 . as for c ) , since @xmath90 and @xmath190 are frchet spaces it is enough to show that @xmath256 is separately continuous . clearly , given @xmath96 , the map @xmath263 is continuous since @xmath217 . on the other hand ,
given @xmath264 in @xmath190 , the hypothesis states that @xmath265 .
lemma [ ltre ] shows that @xmath266 where the identification is given by , namely @xmath267 so that @xmath268 is continuous . finally , we prove d ) .
the definition of @xmath93 and imply . given @xmath269 , by definition @xmath270 and ,
hence , the convolution @xmath271 exists .
furthermore , by assumption @xmath272 . hence , condition [ c3 ] ) of lemma [ rsei ] is satisfied and this implies that @xmath242 , which is . to prove and ,
observe that the reproducing formula implies that @xmath273 and equality that @xmath274 .
furthermore , since @xmath275 , implies @xmath276 and , hence , @xmath277 .
the above result is an adaptation of theorem 2.3 of @xcite .
conditions ( r3 ) and ( r4 ) in @xcite are replaced by and the reproducing property , respectively .
under all the assumptions of proposition [ cuno ] , in particular the conditions of item d ) , the space @xmath255 has a natural topology that makes it a frchet space .
[ topology ] the space @xmath255 is a frchet space with respect to any of the following equivalent topologies : a. the topology induced by the family of semi - norms @xmath278 , where @xmath130 is any fundamental family of semi - norms of @xmath90 ; b. the initial topology induced from the topology of @xmath90 by the restriction of the voice @xmath200 ; c. the final topology induced from the topology of @xmath90 , restricted to @xmath93 , by the fourier transform at @xmath78 . by proposition [ cuno ] , @xmath200 is a bijection from @xmath255 onto @xmath93 whose inverse is the fourier transform at @xmath78 , the initial and final topologies on @xmath255 coincide and they realize @xmath255 as a frchet space ( isomorphic to @xmath93 ) . the equivalence between a ) and b ) is a standard result ( see remark before example 4 ch . 2.11 of @xcite ) . the equivalence between b ) and
c ) follows from the fact that @xmath200 is a bijection . since @xmath93 is a closed subspace of a frchet space , then both @xmath93 and @xmath255 are ( isomorphic ) frchet spaces . in this section
, we further assume that @xmath5 is a _ reproducing _ representation and that the vector @xmath78 is an admissible vector for @xmath5 .
this means that the voice transform @xmath8 maps @xmath6 into @xmath94 and that for all @xmath9 @xmath279 to stress that the voice transform is an isometry of @xmath6 into @xmath94 , we write it with the suffix @xmath280 : @xmath281 recalling that @xmath33 and , we have @xmath282 in the following proposition , some consequences of the assumption that @xmath5 is reproducing are drawn . the results are well known for irreducible representations @xcite and their extensions to non - irreducible representations are taken for granted in many papers .
we provide a proof based on proposition [ cuno ] .
[ tre ] suppose that @xmath5 is a reproducing representation of @xmath0 on @xmath6 and that @xmath7 is an admissible vector .
then : a. for every @xmath283 , the fourier transform of @xmath109 at @xmath78 exists in @xmath6 and for all @xmath9 @xmath284 b. for every @xmath283 the convolution @xmath285 exists and @xmath286 where both sides belong to @xmath287 and , for every @xmath9 , @xmath288 in particular , @xmath289 .
c. the space @xmath290 is a @xmath63-invariant closed subspace of @xmath94 and @xmath291 hence , the voice transform @xmath292 is a unitary map from @xmath6 onto @xmath293 whose inverse is given by the map @xmath294 .
we first prove . by with @xmath295 ,
the convolution @xmath296 exists and is in @xmath287 because @xmath297 by .
furthermore , given @xmath66 , for all @xmath68 @xmath298 integrating with respect to @xmath299 , we obtain @xmath300 to prove the remaining statements , we apply proposition [ cuno ] with @xmath301 and @xmath302 , with the understanding that @xmath303 and @xmath29 are the canonical inclusions , @xmath304 , @xmath305 and @xmath306 .
observe that is satisfied since @xmath307 and , by , the reproducing formula holds for every @xmath9 , regarded as anti - linear form on @xmath6 .
furthermore , by in the appendix with @xmath295 , for all @xmath283 the function @xmath285 is in @xmath287 , taking into account .
in this section , we assume that the representation @xmath5 is reproducing and that the vector @xmath7 is admissible , as in section [ sec : reprod ] . we will construct a coorbit space theory based on the choice of a suitable target space @xmath1 embedded in @xmath31 .
we choose a frchet space @xmath1 with a. a continuous embedding @xmath308 ; b. a continuous representation @xmath102 of @xmath0 acting on @xmath1 such that @xmath309 for all @xmath66 ; c. a continuous involution @xmath99 such that @xmath310 , so that @xmath1 enjoys all the properties of the space @xmath90 in section [ sec : target ] , from which we adopt the notations . in particular , as in , we put @xmath311 the classical theory corresponds to the choice @xmath43 , or a weighted version of it .
the following assumptions are at the root of our construction and are trivially satisfied for @xmath15 .
[ h1 ] the kernel @xmath39 is in @xmath312 and @xmath313 for all @xmath42 , _
i.e _ @xmath314 . [ h2 ] for all
@xmath315 and all @xmath9 we have @xmath316 , _ i.e _ @xmath317 . [ h3 ] the linear space spanned by the orbit @xmath318 is dense in @xmath319 . by proposition
[ uno ] , assumption [ h1 ] implies that for all @xmath42 the convolution @xmath158 exists and @xmath319 is a closed @xmath102-invariant subspace of @xmath1 , so that @xmath320 is a subspace of @xmath321 . assumption [ h3 ] is formulated with a slight abuse of notation , regarding @xmath39 as an element of @xmath1 .
it is a strengthening of assumption [ h1 ] because it is equivalent to the requirement that @xmath39 is actually a cyclic vector for the representation @xmath102 restricted to @xmath319 .
assumptions [ h1 ] and [ h2 ] should be compared with hypotheses ( r2 ) and ( r3 ) of @xcite . in our approach
they are needed to define the test space , as in the classical setting , whereas in @xcite the test space is given a - priori .
we are now in a position to define the space of _ test signals _
, namely @xmath322 we define the _ restricted voice transform _
@xmath323 as the unique map satisfying @xmath324 , that is , for all @xmath325 and @xmath66 we put @xmath326 where @xmath327 is the canonical inclusion .
it is by means of @xmath328 that we topologize @xmath16 : we endow @xmath16 with the initial topology induced by @xmath328 .
as it will be shown in theorem [ tuno ] below , this is just an explicit description of the topology that @xmath16 naturally inherits as coorbit space , because @xmath329 . observe that assumption [ h1 ] implies that @xmath330 , since @xmath331 .
[ tuno ] the space @xmath16 is a frchet space isomorphic to @xmath332 via @xmath292 , and @xmath333 . the canonical embedding @xmath334 is continuous and has dense range .
the transpose @xmath335 is continuous , injective and has dense range .
the representation @xmath5 leaves @xmath16 invariant , its restriction @xmath191 to @xmath16 is a continuous representation of @xmath0 acting on @xmath16 and @xmath78 is a cyclic vector of @xmath191 .
we first prove that @xmath16 is a frchet space .
let @xmath336 and @xmath337 . by the properties i ) , ii ) and iii ) stated at the beginning of this section , we are in the general setting of section [ sec : ext ] .
observe that @xmath338 , @xmath339 and clearly @xmath329 .
furthermore , the fact that @xmath5 is reproducing implies that @xmath292 is injective and , by in proposition [ tre ] , the reproducing formula holds true for all @xmath9 .
hence @xmath340 for all @xmath325 , and we actually get @xmath341
. we can apply corollary [ topology ] because the hypotheses of proposition [ cuno ] that imply it are both satisfied : is just assumption [ h2 ] and , as already noticed , the reproducing property holds for all @xmath325 because @xmath5 is reproducing .
hence @xmath16 is a frchet space and @xmath328 induces a topological linear isomorphism from @xmath16 onto @xmath319 . since @xmath342 , clearly @xmath343
since @xmath16 and @xmath319 are isomorphic , in order to show that @xmath303 is continuous it is enough to prove that @xmath29 is continuous from @xmath319 into @xmath94 . both are frchet spaces , hence it is sufficient to show that @xmath344 has sequentially closed graph .
if @xmath136 is a sequence in @xmath319 converging to @xmath109 in @xmath319 and @xmath345 converges to @xmath138 in @xmath94 , then possibly passing to a subsequence , we can assume that @xmath139 converges for almost all @xmath66 .
hence @xmath346 almost everywhere
. item b ) of proposition [ cuno ] gives that @xmath5 leaves @xmath16 invariant .
since for all @xmath66 and @xmath325 @xmath347 the restriction @xmath191 is a continuous representation on @xmath16 because @xmath102 is a continuous representation on @xmath1 .
the fact that @xmath5 is reproducing implies that @xmath348 is dense in @xmath6 , so that @xmath303 has dense range . finally , since @xmath349 for all @xmath66 , assumption [ h3 ] is another way of saying that @xmath78 is a cyclic vector for @xmath191 .
as for the properties of @xmath350 , corollary 3 chapter ii.6.3 of @xcite shows that @xmath303 is continuous from @xmath351 into @xmath352 .
hence , corollary of proposition 5 chapter ii.6.4 of @xcite gives that @xmath350 is continuous from @xmath353 into @xmath354 and @xmath355 .
finally , corollary 2 chapter ii.6.4 of @xcite shows that since @xmath303 is injective and has dense range , @xmath350 has the same properties . as shown in the above proof , all the hypotheses of corollary [ topology ] are satisfied .
this implies that , whenever a fundamental family @xmath130 of semi - norms of @xmath1 is given , then @xmath356 is a fundamental family of semi - norms of @xmath16 .
this is yet another way to get a direct handle on its topology when a family of seminorms of @xmath1 is known .
we regard the dual @xmath36 of @xmath16 as the space of _ distributions _ and we define the _ extended voice transform _ on it by setting for all @xmath357 @xmath358 the definition works because @xmath16 is @xmath191-invariant , @xmath330 and @xmath191 is a continuous representation .
the following theorem states the main properties of @xmath328 and @xmath200 .
we recall that the contragredient representations @xmath199 and @xmath359 are continuous representations acting on @xmath36 and @xmath360 , respectively , where the dual spaces are endowed with the topology of the convergence on compact subsets ( see proposition 3 chapter viii.2.3 of @xcite ) . furthermore , since @xmath5 is a reproducing representation , proposition [ tre ] ensures that for all @xmath283 the fourier transform of @xmath109 at @xmath78 exists in @xmath6 .
[ lsei ] the restricted voice transform @xmath328 is an injective strict morphism from @xmath16 into @xmath1 with image @xmath319 . for all @xmath315
, we have @xmath361 and , for all @xmath325 , we have @xmath362 furthermore , @xmath328 intertwines @xmath191 and @xmath102 and its transpose @xmath363 is a surjective continuous map , intertwining the representations @xmath359 and @xmath199 .
the extended voice transform @xmath200 intertwines @xmath199 with @xmath63 , is injective and continuous from @xmath36 to @xmath92 , where both spaces are endowed with the topology of compact convergence .
finally , for all @xmath364 , we have @xmath365 by theorem [ tuno ] , @xmath328 induces a topological linear isomorphism from @xmath16 onto @xmath319 , which is a closed subspace of @xmath1 .
corollary 1 , chapter ii.4.2 of @xcite gives that @xmath366 is surjective .
by corollary of proposition 5 , chapter ii.6.4 of @xcite , the map @xmath366 is continuous when both @xmath360 and @xmath36 are equipped with the topology of the simple convergence .
since @xmath5 is reproducing and @xmath367 for all @xmath315 , proposition [ tre ] shows that @xmath368 .
hence , by definition of @xmath16 , @xmath369 and the construction of @xmath328 gives that @xmath370 , where , with slight abuse of notation , @xmath234 is the fourier transform of @xmath371 at @xmath78 .
take now @xmath325 .
since @xmath372 , again proposition [ tre ] yields @xmath373 .
the intertwining property is straightforward : for any @xmath374 @xmath375 injectivity is due to the fact that @xmath78 is cyclic for @xmath191 . to prove that @xmath200 is continuous , fix a compact subset @xmath376 of @xmath0 .
since @xmath377 is continuous , the set @xmath378 is compact in @xmath16 , and @xmath379 is continuous on @xmath36 . finally , take @xmath380 .
then for all @xmath66 @xmath381 we add a remark on the finer topological properties of @xmath200 .
if @xmath382 is a bounded subset of @xmath36 or , equivalently , of @xmath383 , the restriction of @xmath200 to @xmath382 , endowed with the topology of @xmath383 , into @xmath92 , with the topology of the compact convergence , is continuous . indeed , since @xmath16 is a frchet space , then it is barrelled ( corollary of proposition 2 chapter iii.4.2 of @xcite ) . hence @xmath384 ( scolium and definition 2 chapter iii.4.2 of @xcite ) .
proposition 5 chapter iii.3.4 of @xcite implies that on @xmath382 the topology of the simple convergence is equivalent to the topology of precompact subsets .
hence , for any compact subset @xmath385 of @xmath0 , since @xmath377 is continuous , the set @xmath386 is compact in @xmath16 , hence precompact and , by the above reasoning @xmath387 is continuous with respect to the topology of the simple convergence .
the next assumption requires that the reproducing formula holds for any distribution in @xmath36 .
[ h4 ] for all @xmath357 , @xmath388 and @xmath389 .
since the representation @xmath199 leaves @xmath36 invariant and @xmath200 intertwines @xmath199 with @xmath63 , the requirement @xmath390 implies that @xmath271 exists .
furthermore , if @xmath391 is contained in @xmath392 , then @xmath393 holds for all @xmath357 since @xmath394 . in the two propositions that follow , we give sufficient conditions implying assumption [ h4 ] .
[ repr - formula ] assume that @xmath319 is a reflexive space and @xmath390 for all @xmath357 .
then the reproducing formula @xmath395 holds true for all @xmath357 .
since @xmath16 and @xmath319 are isomorphic ( theorem [ tuno ] ) , then also @xmath16 is a reflexive space .
regard @xmath16 as the dual of @xmath36 , which has the property ( gdf ) by proposition 3 chapter 6 .
appendix no.2 of @xcite .
the assumption implies that the map @xmath396 is scalarly integrable from @xmath0 to @xmath397 ,
hence theorem [ gelfand - dunford ] shows that there exists @xmath398 such that @xmath399 by theorem [ tuno ] , @xmath6 is dense in @xmath354 and by @xmath400 , which means that @xmath401 given @xmath68 , by applying the above equality to @xmath402 , we get @xmath403 where the last line is due to the change of variable @xmath404 and the fact that @xmath5 is a unitary representation .
hence the convolution @xmath405 exists and is equal to @xmath240 .
the property @xmath390 for all @xmath357 means that the map @xmath396 is scalarly integrable from @xmath0 to @xmath16 , _
i.e. _ , there exists a linear map @xmath406 such that @xmath407 furthermore , since @xmath6 is continuously embedded in @xmath36 and @xmath5 is a reproducing representation , for all @xmath408 we have @xmath409 by theorem [ tuno ] , the map @xmath350 has a dense image in @xmath383 .
however , @xmath410 is continuous with respect to the weak-@xmath411 topology of @xmath36 if and only if @xmath412 . in the setting of reproducing representations , the requirement that the reproducing formula holds for all distributions
is equivalent to assuming that @xmath413 and , in this case , @xmath410 is precisely @xmath78 .
the hypothesis @xmath412 is precisely property ( r4 ) in @xcite .
furthermore , if @xmath16 is a banach space , as in the classical setting , and if the map @xmath396 is bochner - integrable , then it is scalarly integrable and , clearly , @xmath410 is always in @xmath16 . here is another sufficient condition .
[ suff2 ] assume that @xmath414 and suppose that @xmath415 exists and belongs to @xmath1 .
then @xmath389 for all @xmath357 . by theorem [ tuno ]
, @xmath366 is surjective , so that if @xmath416 , then for any @xmath357 there exists @xmath417 such that @xmath418 .
furthermore , if @xmath415 exists and belongs to @xmath1 , then @xmath419 and , since @xmath420 , the last integral is finite for all @xmath421 . by and
, we have that @xmath422 , hence fubini theorem implies that the convolution @xmath423 exists , and in the appendix shows @xmath424 finally , and give @xmath425 we now fix a banach space @xmath30 , with norm @xmath426 , continuously embedded in @xmath31 and @xmath63-invariant . in order to be consistent with the current literature
, we do not indicate the explicit embedding as we did for the other spaces .
the results in this section hold true under the weaker assumption that @xmath30 is a frchet space .
however , we do not need this generality because the main example that we are interested in is the case when @xmath30 is a weighted @xmath106 space for a fixed value of @xmath153 .
the _ coorbit space _ of @xmath30 is @xmath427 endowed with the norm @xmath428 since @xmath200 is a linear injective map , @xmath429 is clearly a norm .
we will prove below that @xmath430 is in fact a banach space .
just as for the target space @xmath1 , the two basic assumptions for the space @xmath30 may be formulated in terms of kthe duals and have to do with the kernel @xmath39 ( compare assumption [ h5 ] below with assumption [ h1 ] ) and with the image of the voice transform @xmath431 ( compare assumption [ h6 ] below with assumption [ h2 ] ) .
they should also be compared with the corresponding assumptions made in @xcite .
as above , we write @xmath432 [ h5 ] for all @xmath433 , we have @xmath434 , that is , @xmath435 .
[ h6 ] for all @xmath436 and all @xmath325 , we have @xmath437 , _
i.e. _ , @xmath438 . by proposition [ uno ]
applied to @xmath439 , assumption [ h5 ] implies that @xmath440 is a @xmath63-invariant closed subspace of @xmath30 .
furthermore , by proposition [ due ] with @xmath441 , assumption [ h6 ] implies that for all @xmath436 the fourier transform of @xmath109 at @xmath78 exists in @xmath36 .
in the following proposition we list the main properties of @xmath430 .
[ prop6 ] the space @xmath430 is a banach space invariant under the action of the representation @xmath199 .
the extended voice transform is an isometry from @xmath430 onto @xmath440 and its inverse is the fourier transform at @xmath78 .
therefore @xmath442 the proposition is a restatement of proposition [ cuno ] and corollary [ topology ] with @xmath441 and @xmath443 .
the hypothesis is assumption [ h6 ] and the hypothesis in item d ) of proposition [ cuno ] is satisfied by assumption [ h3 ] and assumption [ h4 ] . as in the classical setting
, we have the following canonical identification .
[ coltwo ] the hilbert space @xmath94 satisfies assumptions @xmath444 and @xmath445 , and @xmath446 . since @xmath5 is a reproducing representation , assumptions [ h5 ] , and [ h6 ] are clearly satisfied , and @xmath447 .
take now @xmath448 . by proposition [ prop6 ] @xmath449 .
however , since @xmath450 , by proposition [ tre ] @xmath451 . even though @xmath1 is not a banach space ,
the space @xmath452 is well defined and , under assumption [ h4 ] , corollary [ coltwo ] and the definition of @xmath16 imply that , as in the classical setting , @xmath453 .
the above identification suggests to characterize the space @xmath454 the equality @xmath455 is equivalent to require that @xmath456 for all @xmath42 and @xmath357 , that is , @xmath457 , which is in general stronger than assumption [ h4 ] .
let us compare our approach with the theory developed by j. christensen and g. lafsson in @xcite .
assumptions [ h1]@xmath458[h6 ] ensure that the test space @xmath16 defined by satisfies the properties ( r1)@xmath458(r4 ) , and some of our claims can be directly deduced by the results contained in @xcite ( for example , compare theorem 2.3 of @xcite with our proposition [ prop6 ] ) . in our setting , which is somehow parallel to the classical @xmath22 case , we first introduce the target space @xmath1 , which is independent of the reproducing representation , and then we define the test space @xmath16 as the set of vectors for which the voice transform belongs to @xmath1 .
the introduction of the target space @xmath1 makes our construction closer to the classical approach by h. feichtinger and k. grchenig , and assumptions [ h1 ] , [ h2 ] , [ h3 ] and [ h4 ] involve only the target space @xmath1 without any reference to the model space @xmath30 .
moreover , our proofs mainly rely on the theory of weak integrals _
la dunford - pettis _ , which allows us to state our hypotheses as integrability conditions , rather than a continuity requirement as in @xcite .
assumption [ h4 ] requires that the reproducing formula @xmath459 holds for all @xmath357 .
however , in the proof of proposition [ prop6 ] , the reproducing formula is needed only for the distributions in @xmath430 ( compare with item d ) of proposition [ cuno ] ) .
the following lemma shows some equivalent conditions , weaker than assumption [ h4 ] , under which proposition [ prop6 ] remains true .
[ lnove ] take @xmath1 and @xmath30 such that assumptions @xmath460 and assumptions @xmath461 hold true .
then the following facts are equivalent : a. [ item:1 ] for all @xmath462 , @xmath463 ; b. [ item:2 ] for all @xmath462 , @xmath236 exists and @xmath464 ; c. [ item:3 ] for all @xmath462 , the map @xmath465 is in @xmath15 and @xmath466 d. [ item:4 ] for all @xmath462 , the map @xmath467 is scalarly integrable and its scalar integral is @xmath468 , that is @xmath469 by definition of coorbit space , @xmath470 whenever @xmath471 .
hence [ item:1 ] ) is equivalent to [ item:2 ] ) .
since @xmath430 is @xmath199-invariant , [ item:3 ] ) implies that the map @xmath472 is integrable for all @xmath66 and @xmath473 hence c ) implies item [ e2 ] ) of lemma [ rsei ] .
the converse is also true by evaluation at the identity .
therefore c ) is equivalent to item [ e2 ] ) of lemma [ rsei ] , which provides the equivalence between [ item:2 ] ) and [ item:3 ] ) and shows that [ item:4 ] ) implies [ item:3 ] ) .
assume now that @xmath463 .
proposition [ due ] with @xmath244 gives that @xmath240 satisfies , that @xmath474 exists and @xmath475 .
finally , since @xmath200 is injective by theorem [ lsei ] , we know from item [ c2 ] ) of lemma [ rsei ] that [ item:1 ] ) implies [ item:4 ] ) .
we now examine the dependence of space @xmath16 on the choice of the admisible vector @xmath78 . for this reason , in this section , we write @xmath476 instead of @xmath16 , and accordingly for other choices of admissible vectors .
[ dependency ] suppose that @xmath477 and that for all @xmath478 the map @xmath479 is continuous from @xmath1 into itself , where @xmath480 . if @xmath481 is another admissible vector satisfying assumptions @xmath482 , @xmath483 and @xmath484 , then the test function spaces @xmath485 and @xmath476 coincide as frchet spaces .
furthermore , @xmath486 for any admissible @xmath78 and @xmath487 for all @xmath488 .
let @xmath489 and @xmath66 . since @xmath5 is reproducing and @xmath78 is admissible , @xmath490 where @xmath491 since @xmath492 .
the hypothsis on @xmath1 implies that @xmath493 , so that @xmath494 .
we now prove that the embedding of @xmath476 into @xmath485 is continuous .
fix a semi - norm @xmath495 of @xmath485 , _ i.e _
, fix a semi - norm @xmath496 of @xmath1 such that @xmath497 for all @xmath498 . by with @xmath499 and @xmath500 ,
there exist a constant @xmath128 and a semi - norm @xmath501 of @xmath1 such that @xmath502 where @xmath503 is a semi - norm of @xmath476 .
hence , the embedding is continuous .
interchanging the roles of @xmath78 and @xmath504 , we obtain that @xmath505 with a continuous embedding . finally by in the appendix and , for all @xmath489 , @xmath506 by assumption , so that @xmath507 and , hence , @xmath508 . in the classical framework
, @xmath5 is irreducible and @xmath509 , where @xmath510 is a continuous density satisfying and in the appendix and @xmath511 this last condition implies that @xmath512 so that the hypotheses of the above proposition are satisfied .
however , a stronger result holds true , namely @xmath513 which is the content of lemma 4.2 in @xcite .
note that the irreducibility ensures that , if @xmath514 , then @xmath78 is an admissible vector .
however , if @xmath5 is not irreducible , the above equality does not hold as shown by a counter - example in @xcite .
in this section , we illustrate some examples . they include band - limited functions ( section [ band ] ) , shannon wavelets ( section [ shannon ] ) and _ schrdingerlets _ ( section [ scr ] ) that have inspired our theory . in this section , @xmath517 will denote a continuous function , to be called _ weight _ , satisfying @xmath518 for all @xmath374 . as a consequence
, it also holds that @xmath519 the notion of weight in @xcite is based on the submultiplicative property
. the symmetry can always be satisfied by replacing @xmath510 with @xmath520 .
this requirement is necessary for our development ( see item g ) of theorem [ intersections ] below ) .
condition is explicitly stated in @xcite and , in the classical @xmath15 setting , it is necessary to ensure that the test space is a banach space ( see theorem [ sbanach ] below ) . in the usual irreducible
@xmath22 setting , it is also assumed that the weight satisfies , which is actually incompatible with .
however , is only necessary in order to see that the space of admissible vectors coincides with the test space ( see lemma 4.2 in @xcite ) . in the non irreducible case ,
though , this set - theoretic equality is lost anyhow , as mentioned in the introduction @xcite . for all @xmath521 define the separable banach space @xmath522 with norm @xmath523 and the obvious modifications for @xmath524 .
clearly , the map @xmath525 defined by @xmath526 is a unitary operator .
the following characterization of the kthe dual holds true .
[ 4.1 ] fix @xmath527 and denote by @xmath528 $ ] the dual exponent .
then @xmath529 for all @xmath530 and @xmath531 , set @xmath532 then the map @xmath533 is an isomorphism from @xmath534 onto @xmath535 . under this identification , the transpose @xmath536 is given by @xmath537 for @xmath538 we have @xmath539 if and only if @xmath540 for every @xmath531 , which is equivalent to @xmath541 for every @xmath542 .
this , in turn , happens if and only if @xmath543 , which means that @xmath544 .
hence @xmath545 , the pairing @xmath546 is the pairing between @xmath547 and @xmath515 given in lemma [ ltre ] , and @xmath548 thus the map @xmath533 is an isometry from @xmath534 into @xmath535 and it allows to identify @xmath534 with a closed subspace of @xmath535 .
we now compute the transpose of @xmath549 , taking into account that @xmath550 . for a fixed @xmath551 and
all @xmath531 we have @xmath552 so that , since @xmath510 is positive , @xmath553 . since @xmath549 is unitary , so is @xmath554 and @xmath555 .
[ leftreg ] for all @xmath527 , the left regular representation leaves @xmath515 invariant .
the restriction @xmath102 of @xmath63 to @xmath515 is a continuous representation with @xmath556 for all @xmath66 .
fix @xmath66 . by , for all @xmath531 @xmath557 so that @xmath558 leaves @xmath515 invariant and the norm of the restriction @xmath559 is bounded by @xmath560 .
we now prove that @xmath102 is continuous by applying the concluding remark of section [ reps ] in the appendix . for any compact subset @xmath385 of @xmath0 ,
since @xmath510 is continuous , @xmath561 is bounded and , hence , @xmath562 is equicontinuous .
furthermore , if @xmath563 , the map @xmath564 is clearly continuous from @xmath0 into @xmath565 by the dominated convergence theorem .
the proof is completed by observing that @xmath566 is a dense subset of @xmath515 .
let @xmath567 .
we define the target space as the set @xmath568 with the initial topology , which makes each inclusion @xmath569 continuous , and endow @xmath570 with the final topology , which makes each inclusion @xmath571 continuous . recall that by definition of initial and of final topology , for any topological space @xmath572 , a map @xmath573 is continuous if for all @xmath574 there exists a continuous map @xmath575 such that @xmath576 , and a map @xmath577 is continuous if all @xmath578 there exists a continuous map @xmath579 such that @xmath580 . as for notation ,
given the nature of @xmath581 , the inclusion @xmath582 is set - theoretically tautological because the elements of @xmath581 are ( equivalence classes of ) measurable functions .
however , we keep it to emphasize that the two spaces , @xmath581 and @xmath24 , have different topologies . the following theorem states the main properties of @xmath581 and @xmath583 .
[ tu ] the space @xmath581 is a reflexive frchet space , whose topology is given by the fundamental family of semi - norms @xmath584 .
it is closed under complex conjugation and @xmath99 is continuous .
the canonical inclusion @xmath585 is continuous , the left regular representation @xmath80 leaves @xmath581 invariant and the restriction @xmath102 of @xmath63 to @xmath581 is a continuous representation of @xmath0 on @xmath581 .
the space @xmath583 is a complete reflexive locally convex topological vector space . for each @xmath586 ,
the anti - linear map from @xmath581 into @xmath258 given by @xmath587 is continuous and @xmath588 identifies , as topological vector spaces , the dual of @xmath581 with @xmath583 .
furthermore the kthe dual of @xmath581 is @xmath583 , so that @xmath589 for each @xmath590 , the anti - linear map from @xmath583 to @xmath258 @xmath591 is continuous and @xmath592 identifies , as topological vector spaces , the dual of @xmath583 with @xmath581 . the proof is based on the content of the article @xcite , whose main results are summarized by theorem [ classic ] in the appendix , where @xmath593 and @xmath594 ( in @xcite it is assumed that @xmath595 ) . by definition of initial topology , @xmath581 is a locally convex topological space and @xmath584 is a fundamental family of semi - norms .
clearly , @xmath581 is closed under complex conjugation and is left invariant by @xmath63 .
we show that @xmath102 is a continuos representation .
given @xmath66 , @xmath596 is continuous because @xmath597 . given @xmath590 , the map @xmath598 is continuous from @xmath0 to @xmath581 since such are the maps @xmath599 for all @xmath574 . the proof that complex conjugation is continuous is similar .
define the linear map @xmath600 , @xmath601 .
since @xmath602 , @xmath603 is a bijection whose inverse is given by @xmath604 .
both maps are continuous by definition of initial topology since @xmath605 for all @xmath153 .
hence @xmath603 is a topological isomorphism . by theorem [ classic ]
, we infer that @xmath1 is a reflexive frchet space and , hence , @xmath581 is a reflexive frchet space , too .
define @xmath606 , @xmath607 , which is clearly bijective and whose inverse is @xmath608 . by definition of final topology , both are continuos since for all @xmath578 @xmath609 ( with slight abuse , here @xmath610 denotes the inclusion of @xmath611 into @xmath612 ) . hence @xmath613 is an isomorphism from @xmath612 onto @xmath583 .
therefore , by theorem [ classic ] , @xmath583 is a complete barelled locally convex topological vector space since such is @xmath612 . since @xmath603 is an isomorphism between two frchet spaces , by corollary 5 of chapter iv.4.2 of @xcite
, @xmath614 is an isomorphism from @xmath612 onto @xmath615 explicitly given by @xmath616 where @xmath617 with @xmath618 and @xmath619 , , @xmath620 .
hence , we can identify @xmath615 and @xmath583 as topological vector spaces by means of the map @xmath621 , and the pairing between @xmath583 and @xmath581 is @xmath622 observe that implies @xmath623 for all @xmath145 , so that @xmath624 . furthermore , ensures that @xmath625 if and only if @xmath626 .
+ we are now ready to state the main result of this section .
[ intersections ] take a reproducing representation @xmath5 of @xmath0 acting on the hilbert space @xmath6 and a weight @xmath510 satisfying , and .
choose an admissible vector @xmath7 such that @xmath627 and set @xmath628 then : a. the space @xmath629 is a reflexive frchet space with respect to the topology induced by the family of semi - norms @xmath630 , the canonical inclusion @xmath631 is continuous and with dense range ; b. the representation @xmath5 leaves @xmath629 invariant , its restriction @xmath191 is a continuous representation acting on @xmath629 , and @xmath632 c. if @xmath6 and @xmath633 are endowed with the weak topology , the transpose @xmath634 is continuous , injective , with dense range and satisfies the intertwining @xmath635 d. the restricted voice transform @xmath636 , given by @xmath637 is an injective strict morphism onto the closed subspace @xmath638 and it intertwines @xmath191 and @xmath102 ; e. [ bar ] every @xmath590 , has at @xmath78 a fourier transform in @xmath629 and @xmath639 furthermore , the map @xmath640 is continuous and its restriction to @xmath641 is the inverse of @xmath328 ; f. [ sopra ] every @xmath642 has at @xmath78 a fourier transform in @xmath643 and @xmath644 g. the extended voice transform given by takes values in @xmath583 , it is injective , continuous ( when both spaces are endowed with the strong topology ) and intertwines @xmath199 and @xmath63 ; the range of @xmath200 is the closed subspace @xmath645 and for all @xmath646 and @xmath647 @xmath648 h. [ sotto ] the map @xmath649 is the inverse of @xmath200 and coincides with the restriction of the map @xmath366 to @xmath650 , namely @xmath651 i. @xmath652 .
the fundamental requirement states that @xmath653 , whereas is the reconstruction formula for the distributions in @xmath643 , namely the fact that for any @xmath646 the formula @xmath654 holds for all @xmath647 , where the integral converges since @xmath655 is in @xmath581 by definition of @xmath629 , and @xmath656 is in @xmath657 since the range of @xmath200 is contained in @xmath583 ( for arbitrary target spaces this last property could fail ) . since @xmath658 ,
assumption [ h1 ] is satisfied .
furthermore , since @xmath659 by , also assumption [ h2 ] is satisfied .
the topology induced by the family of semi - norms @xmath630 is the initial topology on @xmath629 induced by the map @xmath328 , as in the proof of corollary [ topology ] .
a. by theorem [ tuno ] which does not depend on assumption [ h3 ] , the space @xmath629 is a frchet space isomorphic to @xmath641 and the canonical inclusion @xmath631 is continuous and with dense range .
furthermore , since @xmath641 is a closed subspace of a reflexive frchet space , both @xmath641 and @xmath629 are reflexive .
b. apply theorem [ tuno ] .
c. apply theorem [ tuno ] for the main statement ; the intertwining property is easily checked .
d. apply theorem [ lsei ] .
e. fix @xmath590 . by @xmath367 and by proposition [ tre ]
there exists @xmath660 such that @xmath661 .
we claim that @xmath371 and @xmath39 are convolvable ( see the appendix for the definition ) and @xmath662 .
it is enough to show that for all @xmath663 , @xmath664 .
indeed , given @xmath66 @xmath665 by .
define @xmath666 , so that @xmath667 . then by assumption @xmath668 , @xmath669 and @xmath670
. hence applies , showing that @xmath671 and @xmath672 are convolvable , @xmath673 and @xmath674 , where @xmath174 is a constant depending on @xmath675 and @xmath39 . hence @xmath676 and @xmath677 therefore @xmath678 . by definition of @xmath629 , @xmath679 and @xmath680 .
the map @xmath681 is continuous by . by theorem [ lsei ] ,
the map @xmath682 is the inverse of @xmath328 .
f. fix @xmath683 . by linearity
we can assume that @xmath684 is in some @xmath685 , so that @xmath686 is in @xmath15 for all @xmath590 .
in particular for all @xmath647 we have @xmath687 because @xmath688 , and by proposition [ due ] there exists @xmath689 . furthermore , recalling that @xmath366 is a linear map from @xmath615 onto @xmath643 , for all @xmath647 we have @xmath690 comparing this equation with the definition of @xmath691 we get @xmath692 and , by , @xmath693 . g. fix @xmath694 . since @xmath695 is surjective , there exists @xmath696 such that @xmath697 , so that @xmath698 . to show that @xmath699 we prove that @xmath684 and @xmath39 are convolvable whenever @xmath683 , their convolution @xmath700 is in @xmath583 and the map @xmath701 is continuous from @xmath583 into @xmath583 . by definition of @xmath583 ,
it is enough to show that , given @xmath574 , for all @xmath702 , @xmath684 and @xmath39 are convolvable , @xmath703 and the map @xmath701 is continuous from @xmath685 into @xmath704 . as above ,
@xmath705 where in the third line we used both and . by assumption @xmath706 ,
@xmath669 where we set @xmath707 , and @xmath670 .
since @xmath708 , gives that @xmath709 and @xmath672 are convolvable , @xmath710 and @xmath711 where @xmath174 is a constant depending on @xmath153 and @xmath39 . hence @xmath712 and @xmath713 this proves the claim .
note that , since @xmath701 is continuous , @xmath650 is closed .
we observe _ en passant _ that @xmath583 is not a frchet space , so that proposition [ uno ] does not apply .
+ we now prove that @xmath714 . since @xmath715 , reasoning as in the proof of item [ bar ] ) , @xmath716 .
furthermore , if @xmath717 , so that @xmath718 , then as above @xmath684 and @xmath125 are convolvable and gives that @xmath719 so that the range of @xmath200 is contained in @xmath720 . from [ sopra ] ) we know that @xmath721 whenever @xmath722 and that @xmath723 , showing that @xmath200 is onto @xmath720 and that the map @xmath724 is the inverse of @xmath200 , as claimed in item [ sotto ] ) .
+ next , we prove .
fix @xmath725 and define the map @xmath726 by @xmath727 for all @xmath694 , @xmath728 and @xmath729 , so that @xmath730 is in @xmath15 .
+ since @xmath629 is a reflexive frchet space , we can regard @xmath629 as the dual of @xmath643 , which has the property ( gdf ) by proposition 3 chapter 6 .
appendix no.2 of @xcite .
the fact that the map @xmath730 is in @xmath15 , means that @xmath229 is scalarly integrable .
theorem [ gelfand - dunford ] shows that its ( scalar ) integral is in @xmath629 , _
i.e. _ that there exists @xmath731 such that @xmath732 with the choice @xmath733 , @xmath734 , gives that @xmath735 .
since this last equality holds true for all @xmath734 and @xmath303 is injective , then @xmath736 and this proves .
furthermore , the reproducing formula implies that @xmath200 is injective .
+ finally , we prove that @xmath200 is continuous .
fix a bounded subset @xmath382 in @xmath581 . by e )
the map @xmath737 is continuous from @xmath581 into @xmath629 .
then @xmath738 is a bounded subset of @xmath629 .
furthermore , given @xmath694 @xmath739 since @xmath740 is a bounded subset of @xmath629 , the map @xmath741 is continuous , hence @xmath200 is such .
the rightmost equality is a consequence of the definition of @xmath742 and the inclusion follows from the fact that @xmath743 .
indeed , for all @xmath66 @xmath744 where @xmath174 is a constant depending on @xmath468 , @xmath745 are suitable numbers in @xmath746 also depending on @xmath468 , and the last bound is a consequence of lemma [ leftreg ] .
h. see the proof of the above item .
i. apply d ) of proposition [ cuno ] with @xmath747 and @xmath748 , taking into account [ bar ] ) .
we summarize the findings in this section in the following theorem which is one of the main results of this paper since it shows that our analysis is indeed applicable .
[ stephan ] if @xmath749 , then assumptions @xmath750 are satisfied for @xmath581 . under the hypothesis ,
assumption [ h1 ] is satisfied , and @xmath659 implies assumption [ h2 ] .
the reconstruction formula clarifies that @xmath78 is a cyclic vector for @xmath191 , which is equivalent to assumption [ h3 ] since @xmath349 and @xmath328 is an injective strict morphism from @xmath629 onto @xmath641 .
finally , with @xmath751 implies that assumption [ h4 ] holds true .
observe that theorem [ stephan ] paves the way for a coorbit space theory with a specific choice of target space , namely @xmath581 .
indeed , if @xmath30 is a banach space continuously embedded in @xmath31 and @xmath63-invariant , and we assume that @xmath752 and that @xmath753 is a subspace of @xmath583 , then assumptions @xmath444 and @xmath445 are satisfied . hence proposition [ prop6 ] holds true , giving rise to a coorbit theory for @xmath30 .
we summarize the general scheme in the following picture . as a toy example
, we consider the space of band - limited signals on the real line . although elementary , and certainly very natural , this case can not be handled by the classical coorbit machinery .
this is somewhat unsatisfactory , because the sinc function is one of the first examples of reproducing kernels which comes to mind .
our theory does handle it , and the natural coorbit spaces that arise are the paley wiener @xmath153-spaces . in this section ,
@xmath0 is the additive group @xmath754 and the haar measure is the lebegue measure @xmath755 .
we denote by @xmath756 the frchet space of rapidly decreasing functions and by @xmath757 the space of tempered distributions .
the fourier transform on @xmath758 and @xmath757 is denoted by @xmath759 .
regarding @xmath760 as a subspace of @xmath757 , we set @xmath761 for any @xmath762 . the representation @xmath5 is the regular representation restricted to the paley
wiener space of functions with band in the fixed compact interval @xmath763 , namely @xmath764 strictly speaking , the elements of @xmath765 are not functions , but equivalence classes of functions . however , in view of the paley wiener
schwartz theorem @xcite , each equivalence class in @xmath765 has a unique representative which is continuous ( in fact smooth ) .
we are thus allowed to identify each class with its smooth representative , and we shall do so .
since the group @xmath766 acts on @xmath765 by translations : @xmath767 on the frequency side , @xmath768 acts on @xmath769 by modulations : @xmath770 this representation is not irreducible : any subset @xmath771 gives a subrepresentation on @xmath772 .
for the reader s convenience , we summarize in the next proposition the main facts that are relevant to our discussion . [ prop : band ] the representation @xmath5 is reproducing and the following facts hold true . a. [ item : band - wavelet ] a vector @xmath773 is admissible if and only if @xmath774 almost everywhere on @xmath775 . in this case , the kernel @xmath39 is @xmath776 where @xmath777 is the characteristic function on @xmath775 .
b. [ item : band - voice ] let @xmath78 be an admissible vector .
then @xmath778 for every @xmath779 if and only if the corresponding voice transform @xmath292 is the inclusion @xmath780 c. if @xmath781 $ ] is a symmetric interval , then the kernel is the sinc function @xmath782 where @xmath783 . the fact that @xmath5 is reproducing follows from item a ) .
a. applying the plancherel identity , we can compute the voice transform as @xmath784 whose squared norm , again by plancherel , is @xmath785 on the other hand , plancherel also entails @xmath786 therefore , @xmath78 is admissible if and only if @xmath787 for almost every @xmath788 .
since @xmath775 is compact , vectors @xmath789 satisfying the above condition clearly exist and , hence , @xmath5 is reproducing .
if @xmath78 is admissible , using we obtain @xmath790 .
b. suppose that @xmath791 .
then , in view of item we have @xmath792 , and equality gives @xmath793 for every @xmath794 , so that @xmath795 is the natural inclusion .
conversely , if this is the case , then @xmath796 . c. if @xmath797 $ ] , from it follows @xmath798 from now on we set @xmath799 $ ] and @xmath800 clearly , @xmath39 is not in @xmath801 , but it belongs to @xmath802 for every @xmath803 .
we thus choose the weight @xmath595 and take @xmath804 as target space to construct coorbits ( recall that @xmath567 ) . for @xmath805
, we define the paley
wiener @xmath153-spaces @xmath806 recall that the fourier transform maps @xmath106 to @xmath807 for @xmath808 , while for @xmath809 we get distributions that in general are not functions @xcite .
the spaces @xmath810 are usually defined in the literature as the spaces of the entire functions of fixed exponential type whose restriction to the real line is @xmath153-integrable @xcite .
this definition is equivalent to ours since a paley
wiener theorem holds for all @xmath805 .
in particular , all these functions are indefinitely differentiable on @xmath754 . moreover , if @xmath811 with @xmath155 , then @xmath812 as @xmath813 , hence @xmath814 consequently , the paley
wiener spaces are nested and increase with @xmath153 : @xmath815 we are going to identify our coorbit spaces as paley wiener spaces . to show this , we shall repeatedly make use of the following fact .
[ lemma : band ] there exists a family of functions @xmath816
satisfying a. @xmath817 in @xmath818 for every @xmath819 ; b. @xmath820 } \leqslant \widehat{g_\varepsilon } \leqslant \chi_{[-\omega-\varepsilon,\omega+\varepsilon ] } $ ] ; c. @xmath821 . such that for all @xmath822 , with @xmath823 , in @xmath824 @xmath825 take @xmath826 . by young s inequality
we know that @xmath827 for some @xmath828 , so that @xmath829 and the left hand side is the fourier transform of a tempered distribution .
similarly , @xmath830 because @xmath831 , and on the right hand side we also have fourier transforms of tempered distributions .
take @xmath832 , \quad \widehat{h } \geqslant 0 , \quad \int_{\widehat\r } \widehat{h}(\xi ) d\xi = 1,\ ] ] and then consider the corresponding approximate identity @xmath833 defined by the dilations of @xmath834 @xmath835 so that @xmath836 , and define @xmath837 . since both factors are @xmath22-functions , the convolution theorem gives that @xmath838 . a classical result ,
see corollary 3.4 of @xcite , shows that @xmath839 in @xmath840 for every @xmath841 and @xmath842 . moreover , @xmath843 + \omega = [ -\omega-\varepsilon,\omega+\varepsilon ] .\ ] ] expanding the convolution , we get @xmath844 by a change of variable .
notice here that if @xmath845 , then @xmath846 whenever @xmath847 , which means that @xmath848 \supseteq { \operatorname{supp}(\widehat{h } ) } $ ] .
it follows that @xmath849 for every @xmath850 $ ] .
the derivative of @xmath851 is @xmath852 , and @xmath853 .
a change of variable then yields @xmath854 } { \lvert{\partial\widehat{h}(\tau)}\rvert } d\tau \leqslant 2\sup({\lvert{\partial\widehat{h}}\rvert } ) \varepsilon^{-1 } .\ ] ] finally , let @xmath855 . by dominated convergence , for all @xmath779 @xmath856 so that @xmath857 and at the same time @xmath858 for any @xmath859
. therefore @xmath860 in @xmath840 , by dominated convergence .
young s inequality implies now that @xmath861 in some @xmath862 , hence as tempered distributions .
therefore @xmath863 by the continuity of the fourier transform and an application of the convolution theorem , because @xmath864 and @xmath865 ( see theorem xv , ch .
vii in @xcite ) .
we are now ready to state the characterization of the natural coorbit spaces relative to band limited functions .
let @xmath781 $ ] and take @xmath866 .
the space of test functions is @xmath867 and the space of distributions is @xmath868 the extended voice transform is the inclusion @xmath869 and the coorbits of the @xmath106 spaces are @xmath870 since @xmath292 is the inclusion by item [ item : band - voice ] ) of proposition [ prop : band ] , we have @xmath871 in order to describe @xmath36 , we first observe that @xmath872 indeed , since @xmath595 , by item @xmath873 , so that @xmath874 whenever @xmath875 . furthermore , item [ sotto ] ) of theorem [ intersections ] implies that @xmath876 establishes a linear isomorphism from @xmath877 to @xmath36 . in particular
, @xmath36 is the range @xmath878 .
since @xmath879 is the inclusion , the transpose map @xmath880 is simply the restriction on the subspace @xmath16 .
therefore , we can explicitly represent @xmath36 as the space of anti - linear functionals @xmath881 the two spaces are thus canonically identified and , with this identification , @xmath200 is the inclusion @xmath882 .
we next prove that @xmath883 for all @xmath574 .
let @xmath884 and @xmath138 be a smooth function with compact support contained in @xmath885 .
then , by lemma [ lemma : band ] , we have @xmath886 for some @xmath851 with @xmath887 $ ] .
since @xmath888 \cap { \operatorname{supp}(\varphi ) } = \emptyset $ ] for @xmath889 small enough , the limit is zero .
this means that @xmath890 , that is @xmath811 .
conversely , let @xmath811 .
we shall prove that @xmath891 , whence @xmath892 and @xmath884 .
thanks to formula iv at page 111 in @xcite , there exists a continuous function @xmath893 , periodic on @xmath775 such that @xmath894 \operatorname{e}^{2\pi ix\xi } d\xi.\ ] ] hence @xmath895 so that , in @xmath896 , @xmath897 this tempered distribution acts on any function @xmath898 by @xmath899 d\xi .
\label{eq : temp } \end{aligned}\ ] ] on the other hand , we know from lemma [ lemma : band ] that @xmath900 where the limit is taken in @xmath901 .
compute now @xmath902 \widehat{g_\varepsilon } d\xi \\ & + \frac{i}{2\pi } \int_\omega ( \psi(\omega ) - \psi(\xi ) ) \varphi(\xi ) \partial \widehat{g_\varepsilon}(\xi ) d\xi .
\end{aligned}\ ] ] by ii ) of lemma [ lemma : band ] , @xmath903 pointwise , then the limit of the first integral is precisely .
it remains to verify that the last integral tends to zero .
notice that it vanishes on @xmath904 $ ] because , by lemma [ lemma : band ] , @xmath905 , hence @xmath906 . the rest of the integral is dominated by @xmath907 { \lvert{\varphi}\rvert}_\infty { \lvert{\partial \widehat{g_\varepsilon}}\rvert}_\infty \ \varepsilon \\
\lesssim & \sup_{-\omega\leqslant\xi\leqslant-\omega+\varepsilon}{\lvert{\psi(\omega)-\psi(\xi)}\rvert } + \sup_{\omega-\varepsilon\leqslant\xi\leqslant\omega}{\lvert{\psi(\omega)-\psi(\xi)}\rvert } , \end{aligned}\ ] ] thanks to iii ) of lemma [ lemma : band ] . but
this tends to zero as @xmath908 , because @xmath909 .
we have proven that @xmath883 for all @xmath153 .
thus we finally obtain @xmath910 1truecm we consider now the special case of a non integrable kernel for the wavelet representation of the affine group on @xmath760 . it is well known that this representation is reproducing and admits admissible vectors whose kernel is integrable @xcite .
the resulting coorbit spaces are completely understood as homogeneous besov spaces @xcite .
however , there are admissible vectors whose kernel is not integrable , as for example the shannon wavelet , which provides another example for which our theory applies . at first sight , this result might look surprising .
indeed , in @xcite , h. feichtinger and k. grchenig claim that any band limited function whose fourier transform has compact support bounded away from zero leads to an integrable kernel .
however , a careful inspection of the proof reveals that the additional assumption that the fourier transform is continuous is implicitly needed .
let @xmath911 be the connected component of the affine group with left haar measure @xmath912 .
the wavelet representation @xmath5 acts on @xmath913 by dilations and translations : @xmath914 the ( real ) shannon wavelet is defined as @xmath915}({\lvert{\xi}\rvert } ) = \chi_{[-1/2,1/2]}(\xi ) - \chi_{[-1/4,1/4]}(\xi ) , \ ] ] that is @xmath916 it is easily seen that @xmath917 , but @xmath918 for all @xmath803 .
we now prove that the corresponding kernel has the same behavior .
the kernel @xmath33 associated with the shannon wavelet @xmath919 is in @xmath920 for all @xmath803 , but it is not in @xmath15 . since @xmath78 is real and even , the voice transform @xmath795 is @xmath921 the shannon kernel is thus @xmath922 since @xmath78 is admissible , @xmath923 and , by fubini theorem , @xmath924 for almost every @xmath925 .
then , by the convolution theorem for @xmath4-functions @xmath926 \cap [ 1/4a,1/2a]}({\lvert{\beta}\rvert } ) .\ ] ] it follows that : a. [ enum : shannon-0 ] @xmath927 only if @xmath928 ; b. [ enum : shannon - f ] if @xmath929 the fourier transform of @xmath930 is a non - zero characteristic function . by [ enum : shannon - f ] ) , for almost all @xmath929 the function @xmath931 can not be in @xmath801 ,
otherwise its fourier transform would be continuous . hence @xmath932 .
let us show that @xmath933 for all @xmath934 . from [ enum : shannon-0 ] ) we have @xmath935 recall that @xmath936 for all @xmath803 , so that the same holds for @xmath937 . by young s inequality for the unimodular group @xmath938 , we can estimate the inner norm and obtain @xmath939 where @xmath153 and @xmath675 are such that @xmath940 .
this integral is finite , because the function @xmath941 is continuous and the interval @xmath942 $ ] is compact . a shannon wavelet coorbit theory
can thus be implemented taking voices in the target space @xmath943 , but not in @xmath15 . in this section ,
we illustrate the example that has motivated the search for a full coorbit theory in which one encounters reproducing kernels that do enjoy nice integrability properties but are not necessarily in @xmath15 .
the main feature of this example is that once the admissibility conditions are worked out , it is relatively easy to exhibit kernels in @xmath944 but hard to find a kernel in @xmath15 .
this example has shown up in the classification of reproducing triangular subgroups of @xmath20 , which was recently achieved in @xcite .
we shall be concerned with the three - dimensional group generated by rotations , dilations and flows of two - dimensional signals , in a sense to be made precise below .
the group acts on functions via radial affine transformations , and the associated voice transform can thus be seen as a fourier series of one - dimensional wavelets .
this representation is highly reducible , and reproducing .
the group @xmath0 is the direct product of the ( connected component of ) affine group of the line with the unit circle @xmath945 and its elements are parametrized by @xmath946 with @xmath947 , @xmath948 and @xmath949 .
a left haar measure is @xmath950 notice that @xmath0 is not unimodular and has modular function @xmath951 .
the representation @xmath5 that we are going to define acts on @xmath952 , endowed with the tensor product of the lebesgue measure and the normalized haar measure on @xmath953 .
the action is @xmath954 since @xmath955 , @xmath5 is simply the tensor product @xmath956 where @xmath510 is the wavelet representation of the affine group @xmath957 and @xmath63 is the left regular representation of @xmath953 on @xmath958 , namely @xmath959 in what follows , we denote by @xmath960 the unitary fourier transform from @xmath760 onto @xmath961 , which is also regarded as a unitary map from @xmath962 onto @xmath963 .
furthermore , we denote by @xmath964 the unitary fourier transform from @xmath958 onto @xmath965 , which is also regarded as a unitary map from @xmath962 onto @xmath966 .
explicitly , if @xmath967 , then @xmath968 where @xmath969 .
the partial fourier transform @xmath970 of any @xmath971 can be identified with the sequence of functions @xmath972 in @xmath760 , where @xmath973 . hence @xmath974 to simplify the computations , we restrict the representation @xmath5 to the closed subspace @xmath975 , so that the wavelet representation @xmath510 acts irreducibly on @xmath976 .
given a vector @xmath977 , we denote by @xmath8 the voice transform corresponding to the representation @xmath5 of @xmath0 and the analyzing vector @xmath78 , namely @xmath978 and by @xmath979 the voice transform corresponding to the representation @xmath510 of the affine group and the analyzing vector @xmath980 , _ i.e. _ @xmath981 we use the unitary operator @xmath982 to obtain an intermediate equivalent version of @xmath5 , denoted @xmath983 , acting on @xmath984 .
this is defined via the intertwining @xmath985 for every @xmath986 .
the analytic expression of @xmath983 is immediately computed to be @xmath987 whereas from the structural point of view it may be written as @xmath988 where @xmath989 .
the group @xmath0 can be realized as the triangular subgroup of @xmath20 consisting of the matrices @xmath990 thus , @xmath0 may also be seen as the semidirect product @xmath991 , where the homogeneous factor @xmath992 acts on the normal subgroup @xmath754 by isotropic dilations .
we shall not distinguish between @xmath953 and @xmath993 and write rotations as @xmath994 we show below that @xmath5 is equivalent to the metaplectic representation @xmath995 as restricted to the above group , defined in the frequency domain by @xmath996 the space - domain version of this representation explains the reason of the name _
schrdingerlets_. denote by @xmath997 the representation obtained by conjugating @xmath995 with the fourier transform , namely @xmath998 we now interpret @xmath947 as a time parameter and look at the evolution _ flow _ of @xmath999 @xmath1000
it is then straightforward to verify that the flow @xmath1001 satisfies the schrdinger equation @xmath1002 where @xmath53 is the spacial laplacian @xmath1003 it is in this sense that the group is generated by ( dilations , rotations and ) flows .
we now prove the equivalence . the unitary map @xmath1004 , defined by @xmath1005 intertwines @xmath995 with @xmath1006 because , for @xmath1007 , we have on the one hand @xmath1008 and on the other hand @xmath1009 as claimed . in conclusion , since @xmath5 and @xmath1006 are equivalent , so are @xmath5 and @xmath995 .
we point out that @xmath229 is simply the change from rectangular to polar - like coordinates @xmath1010 , together with the appropriate @xmath4-normalization . since the wavelet representation is irreducible , while @xmath80 completely reduces to @xmath1011 , where each function @xmath1012 is regarded as a character of @xmath953 , then @xmath1013 which expresses @xmath5 as a sum of irreducibles .
this allows us to view the voice transform of @xmath5 as a fourier series of one - dimensional wavelet transforms , as clarified in the next proposition .
let @xmath1014 .
the voice transform @xmath8 associated with @xmath5 and @xmath78 admits the trigonometric expansion @xmath1015 where the series converges pointwise for all @xmath1016 and where @xmath1017 since @xmath964 is a unitary map , for all @xmath1018 @xmath1019 where the third line is due to the fact that the action of @xmath1020 on @xmath965 is the multiplication operator by the sequence @xmath1021 . for fixed @xmath1022 ,
the function @xmath1023 is continuous and , hence , integrable on @xmath953 .
therefore , by de la valle poussin theorem , ( see iii ) of theorem 11.3 in @xcite ) , we obtain .
the fourier expansion shows how to construct admissible vectors as series of wavelets .
this result has been originally obtained in @xcite and in general setting in @xcite . for the reader
s convenience , we give here a more direct proof .
the representation @xmath5 is reproducing and a vector @xmath1024 is admissible for @xmath5 if and only if for all @xmath1025 @xmath1026 furthermore , given a sequence @xmath1027 with @xmath1028 , we have @xmath1029 if @xmath78 is an admissible vector , the voice transform from @xmath6 into @xmath1030 is @xmath1031 and the series converges also in @xmath94 .
admissibility of @xmath78 means that @xmath1032 must hold for all @xmath9 .
fix @xmath1033 and choose @xmath1034 with @xmath1035 . by , when computing the norm , the integral on the circle is equal to @xmath1036 , whereas the integration on @xmath1037 provides the classical admissibility condition , namely caldern s equations .
conversely , suppose true for every @xmath1033 .
fix @xmath9 . given @xmath1038 , implies that the function @xmath1039 is in @xmath958 if and only if the sequence @xmath1040 is in @xmath965 and , under this assumption , fubini theorem yields @xmath1041 because by for each @xmath1042 the voice @xmath1043 is an isometry from @xmath1044 into @xmath1045 .
the last equality is due to .
equation is a consequence of and the fact that @xmath960 is unitary . to prove that @xmath5 is reproducing , it is enough to show
there exists a sequence @xmath1046 in @xmath1047 satisfying both and .
fix @xmath1048 satisfying , _
i.e. _ , an admissible vector for the wavelet representation @xmath510 . for all @xmath1049 define the functions @xmath1050 as @xmath1051 where @xmath1052 and @xmath1053 .
since @xmath1054 satisfies , so do all the functions @xmath980 .
further , @xmath1055 so that by the vector @xmath1056 is in @xmath6 and is admissible for @xmath5 . finally , we prove . by the series @xmath1057 converges in @xmath94 to @xmath1058 .
now we come to the integrability question .
the idea is based on the very simple observation that caldern s equation is invariant under dilations .
there exist admissible vectors @xmath1059 whose kernel @xmath1060 belongs to @xmath1061 but not to @xmath15 .
define @xmath78 as in the second part of the proof of the above proposition . using we write @xmath1062 where @xmath1063 and the series converges both in @xmath94 and pointwise . by a simple change of variable
, we get that @xmath1064 .
therefore for any @xmath574 @xmath1065 in order to construct @xmath78 , it is therefore sufficient to take a positive sequence for which @xmath1066 converges for every @xmath1067 $ ] .
we now prove that the kernel is not in @xmath15 . by contradiction , assume that @xmath40 .
fubini s theorem implies that for almost all @xmath1068 the function @xmath1069 is in @xmath1070 .
hence , regarding @xmath1071 as an abelian group , its fourier transform @xmath1072 is in @xmath1073 . by
, it holds that @xmath1074 where @xmath1075 is the fourier transform of the function @xmath1076 , which is in @xmath801 by fubini s theorem .
fix @xmath1077 and set @xmath1078 in the above equality
. then @xmath1079 because @xmath1080 .
hence , by injectivity of the fourier transform , @xmath1081 for almost all @xmath1082 . since the above equality holds for almost all @xmath1083 , we get that @xmath1084 , which is a contradiction .
in this section , we apply our machinery and show that the standard setup of coorbit theory makes sense without assuming that the representation @xmath5 is irreducible , because it corresponds to the case arising from the classical choice @xmath1085 .
the fact that irreducibility is a somewhat redundant assumption has been perhaps known to some extent , but it is not easy to pin down precise statements in the literature .
theorem [ nonirreducible ] below contains a summary of the most relevant facts .
it is perhaps worthwhile observing that the present case is structurally different from the case discussed in section [ mainexample ] because @xmath1086 is not a reflexive space . throughout this section ,
we fix a continuous function @xmath517 satisfying the following assumptions ( see @xcite ) : @xmath1087 for all @xmath374 .
we choose as target space @xmath1 the banach space @xmath1088 and denote by @xmath29 the canonical inclusion into @xmath1089 .
since @xmath29 is canonical , we do not write it explicitly , especially because it would conflict with the current literature , where no explicit mention of the embedding is ever made . since lemma [ 4.1 ] and lemma [ leftreg ]
do not depend on assumption , @xmath1090 the left regular representation @xmath63 leaves @xmath1086 invariant , and the restriction @xmath102 of @xmath63 to @xmath1086 is a continuous representation that satisflies @xmath1091 we assume that there exists an admissible vector @xmath977 whose voice transform @xmath1092 is in @xmath1088 and construct the corresponding reservoir @xmath629 of test functions .
we are in a position of stating the main properties of the standard setup , without the assumption that the representation is irreducible .
[ nonirreducible ] take a reproducing representation @xmath5 of @xmath0 acting on the hilbert space @xmath6 and a weight @xmath510 satisfying and .
choose an admissible vector @xmath7 such that @xmath1093 and set @xmath1094 a. the space @xmath1095 is a banach space and the canonical inclusion @xmath1096 is continuous and with dense range .
b. the representation @xmath5 leaves @xmath1095 invariant , its restriction @xmath191 is a continuous representation acting on @xmath1095 , the operator norms satisfy @xmath1097 for all @xmath66 , and @xmath1098 c. endowing @xmath6 with the weak topology and @xmath36 with the weak-@xmath411 topology , the transpose mapping @xmath1099 is continuous , injective and with dense range , and satisfies the intertwining @xmath1100 d. the restricted voice transform @xmath1101 is an isometry intertwining @xmath191 and @xmath63 and its range is the closed subspace @xmath1102 e. for all @xmath1103 , the fourier transform of @xmath109 at @xmath78 exists in @xmath629 and satisfies @xmath1104 furthermore , the map @xmath1105 is continuous and its restriction to @xmath1106 is the inverse of @xmath328 .
f. the extended voice transform @xmath1107 is injective , continuous ( when both spaces are endowed with the topology of the compact convergence ) and intertwines @xmath199 and @xmath63 .
the range of @xmath200 is the closed subspace @xmath1108 for all @xmath694 and @xmath725 @xmath1109 g. for all @xmath1110 the fourier transform of @xmath684 exists at @xmath78 in @xmath643 and satisfies @xmath1111 h. the map @xmath1112 is the inverse of @xmath200 and coincides with the restriction of the map @xmath366 to @xmath1113 , namely @xmath1114 i. @xmath1115 . since @xmath1116 and @xmath1117 for all @xmath9 , assumptions [ h1 ] and [ h2 ] are satisfied .
a. by the first part of theorem [ tuno ] which does not depend on assumption [ h3 ] .
the space @xmath16 is a banach space because @xmath1118 is such .
b. apply theorem [ tuno ] .
moreover , by @xmath1119 c. apply theorem [ tuno ] .
d. apply theorem [ lsei ] .
e. fix @xmath1103 and set @xmath1120 , @xmath1121 .
we show that @xmath229 is bochner - integrable with respect to @xmath52 .
the map @xmath229 is @xmath52-measurable since @xmath1122 and @xmath1123 is continuous from @xmath0 into @xmath1088 , and , by item b ) , @xmath1124 which is in @xmath15 since @xmath1125 .
set @xmath1126 clearly , for all @xmath1127 @xmath1128 hence @xmath1129 satisfies and we can identify @xmath234 with @xmath1129 .
so @xmath1130 .
the fact that @xmath1131 follows from with @xmath1132 and @xmath1133 .
the fact that @xmath294 is the inverse of @xmath328 follows from in proposition [ cuno ] .
f. we first prove that @xmath1134 .
take @xmath1135 . for all @xmath66 , by @xmath1136 so that @xmath1137 is bounded and continuous .
we now prove the reconstruction formula .
fix @xmath1138 and define the map @xmath1139 by @xmath1140 .
we show that it is bochner - integrable with respect to @xmath52 .
it is continuous since both @xmath1141 and @xmath1142 are continuous , and @xmath1143 which is in @xmath15 by definition of @xmath1095 .
hence there exists @xmath1144 such that @xmath1145 for all @xmath1146 we have @xmath1147 and @xmath1148 that is , @xmath1149 .
hence @xmath1150 where the integral is a bochner integral in @xmath629 .
take @xmath1151 .
then for all @xmath1138 @xmath1152 which proves the reconstruction formula .
this , in turn , implies that @xmath200 is injective .
g. apply item a ) of proposition [ cuno ] with @xmath1133 and @xmath1153 .
b ) and d ) of proposition [ cuno ] with @xmath1153 and @xmath1133 show that the range of @xmath200 is the closed subspace @xmath1154 and that the inverse of @xmath200 is @xmath1155 .
i. since @xmath1156 and @xmath1157 for all @xmath1127 , it follows that @xmath1158 .
furthermore , proposition [ cuno ] with @xmath1132 and @xmath1133 gives that @xmath1159 item d ) of this theorem finally implies that @xmath1160 . as shown in the previous proof , assumptions [ h1 ] and [ h2 ] are satisfied .
the reconstruction formula makes clear that @xmath78 is a cyclic vector for @xmath191 , which is equivalent to assumption [ h3 ] because @xmath349 and @xmath328 is an isometry from @xmath629 onto @xmath1106 .
furthermore , with @xmath751 implies that also assumption [ h4 ] holds true . from now until the end of this section we choose a banach space @xmath30 with a continuous embedding @xmath1161 , denoted @xmath1162 , and with a continuous involution @xmath99 .
we further suppose that there are two continuous representations @xmath102 and @xmath1163 of @xmath0 on @xmath30 satisfying a. for all @xmath66 and all @xmath433 @xmath1164 b. for all @xmath433 and almost every @xmath66 , @xmath1165 [ prop4 ] assume that @xmath30 is a banach space with a continuous representation @xmath1163 for which there exists a continuous embedding @xmath1161 , denoted @xmath1166 , such that , for all @xmath66 , all @xmath433 and almost every @xmath68 it holds that @xmath1167 .
suppose that @xmath1168 is such that for all @xmath433 @xmath1169 then @xmath371 and @xmath125 are convolvable , there exists @xmath1170 satisfying @xmath1171 and @xmath1172 the proof is closely related to the proof of proposition 6 chapter viii.4.2 of @xcite .
fix @xmath433 and @xmath1168 and set @xmath1173 we claim that @xmath229 is @xmath52-integrable in the bochner sense .
since @xmath1163 is a continuous representation , the map @xmath1174 is continuous from @xmath0 to @xmath30 and hence it is @xmath52-measurable .
since @xmath1175 , so is @xmath1176 , and hence @xmath229 is @xmath52-measurable .
furthermore , @xmath1177 is @xmath52-integrable by assumption .
hence @xmath229 is @xmath52-integrable .
define @xmath1178 which clearly satisfies @xmath1179 recall that @xmath1180 and take @xmath1181 .
then by with @xmath1182 ( @xmath30 is normed ) @xmath1183 which is finite by assumption . by fubini theorem , the function @xmath1184 is in @xmath1185 and , hence , there exists a negligible set @xmath1186 such that for all @xmath1187 the function @xmath1188 is in @xmath15 .
put @xmath1189 . by the change of variable @xmath1190 , for all @xmath1191 , the function @xmath1192 is in @xmath15 .
take a countable family @xmath1193 such that @xmath1194 and set @xmath1195
. then @xmath179 is negligible and for all @xmath1196 the map @xmath1197 is integrable .
hence , for all @xmath1181 , fubini theorem gives @xmath1198 where @xmath1199 denotes the duality between @xmath30 and @xmath1200 introduced in . by the change of variable @xmath1201 in the inner integral , we get
@xmath1202 this means that @xmath371 and @xmath125 are convolvable , @xmath1203 and , by , the inequality @xmath1204 holds true .
[ myclosed ] take a weight @xmath510 such that @xmath1205 for all @xmath66 .
then : a. for all @xmath433 and @xmath1206 , @xmath371 and @xmath125 are convolvable , there exists @xmath1170 such that @xmath1207 , the map @xmath1208 is continuous and @xmath1209 .
b. the set @xmath1210 is a closed @xmath102-invariant subspace of @xmath30 . item a ) follows from proposition [ prop4 ] , observing that is satisfied and @xmath1211 as for b ) , since @xmath30 is a metrizable topological space , it is sufficient to prove that @xmath440 is sequentially closed .
take a sequence @xmath136 in @xmath440 converging to @xmath433 . possibly passing to a subsequence
, we can assume that there exists a negligible set @xmath179 such that for all @xmath180 the sequence @xmath139 converges to @xmath112 .
furthermore , possibly changing @xmath179 , we can also assume that , for all @xmath1042 and @xmath180 , @xmath1212 .
since @xmath1213 and @xmath29 are continuous , @xmath1214 converges to @xmath1215 in @xmath1216 .
hence , by lemma [ luno ] in the appendix , possibly passing again to a subsequence and again redefining @xmath179 , we can also assume that for all @xmath180 @xmath1217 . then @xmath1218 so that @xmath1219 in @xmath31 , that is @xmath436 . finally , given @xmath66 and @xmath436 , by , @xmath1220 which means that @xmath1221 .
we apply the above corollary with the choice @xmath1222 , which is in @xmath1088 by assumption , together with @xmath1223 .
notice that , although assumption [ h5 ] is not satisfied , b ) of corollary [ myclosed ] guarantees that @xmath1224 is a closed subspace of @xmath30 .
furthermore we assume that @xmath1225 since by construction @xmath1226 for all @xmath325 and @xmath1227 , implies assumption [ h6 ] .
hence we can define @xmath1228 the inclusion is satisfied by all the classical banach spaces considered in @xcite .
this fact is shown in the proof of proposition 4.3 .
the space @xmath430 is a @xmath5-invariant banach space and the restriction of @xmath200 to @xmath430 is an isometry from @xmath430 onto @xmath440 . for all @xmath1229 ,
@xmath691 exists in @xmath633 , it actually belongs to @xmath430 and satisfies @xmath1230 apply proposition [ cuno ] .
1truecm + take a continuous function @xmath517 satisfying only .
take a square - integrable representation @xmath5 acting on @xmath6 and fix an admissible vector @xmath7 such that @xmath1231 and set @xmath1232 [ sbanach ] the space @xmath1095 is a banach space if and only if @xmath1233 assume that @xmath1234 .
we can always suppose that @xmath1235 so that holds true .
indeed , if @xmath1236 , we redefine @xmath510 as @xmath1237 , so that @xmath1238 and @xmath1237 satisfies . since @xmath1239 with equivalent norms , clearly the fact that @xmath1095 is a banach space does not depend on the choice of @xmath1240 .
item a ) of theorem [ nonirreducible ] states that @xmath1095 is a banach space .
assume that @xmath1095 is a banach space .
define @xmath1241 as the vector space @xmath1095 with the norm @xmath1242 we claim that @xmath1241 is complete . take a cauchy sequence @xmath1243 with respect to @xmath1244 . by construction , it is a cauchy sequence also with respect to both @xmath1245 and @xmath1246 . since @xmath1095 and @xmath6 are complete , there exist @xmath1247 and @xmath9 such that @xmath1248 since the voice trasform is an isometry both from @xmath6 into @xmath94 and from @xmath1095 into @xmath1086 , the sequence @xmath1249 converges to @xmath12 in @xmath94 and to @xmath1250 in @xmath1086 .
hence , possibly passing to a subsequence , @xmath1249 converges almost everywhere to @xmath12 and to @xmath1250 .
since @xmath602 , @xmath1251 almost everywhere and , hence , @xmath1252 by the injectivity of @xmath8 , so that @xmath1253 .
furthermore @xmath1254 hence @xmath1241 is complete and the natural inclusion @xmath1255 is clearly continuous and bijective .
since @xmath1095 is a banach space , the open mapping theorem implies that the inverse is also continuous , so that there is a constant @xmath1256 such that @xmath1257 as usual , for all @xmath66 and @xmath1127 @xmath1258 then , if @xmath1259 , for all @xmath66 @xmath1260 taking the infimum over @xmath0 , we get @xmath1261
the set @xmath31 is a metrizable complete topological vector space , and @xmath287 is dense in @xmath31 ( propositions 19 and 20 chapter iv.5.11 of @xcite ) . since @xmath0 is second countable , @xmath287 is separable ( with respect to the convergence on compact subsets , hence with respect to the convergence in measure ) and @xmath31 is separable , too .
furthermore , if @xmath108 is a sequence converging to @xmath109 in @xmath31 , then there exist a subsequence @xmath110 and a negligible set @xmath1262 such that @xmath1263 ( corollary of proposition 19 chapter iv.5.11 of @xcite ) .
if @xmath0 is compact @xmath1264 is a separable banach space .
otherwise , a saturated fundamental system of semi - norms is given as follows ( recall that a family is saturated if the maximum of any finite set of seminorms is in the family ) .
since @xmath0 is second countable , take a countable increasing family @xmath1265 of compact subsets of @xmath0 such that @xmath1266 and @xmath1267 .
for all @xmath1268 put @xmath1269 then , for each compact set @xmath385 there exists @xmath1268 such that @xmath1270 and @xmath1271 with the induced topology , @xmath60 is complete ( see ex .
31 chapter v.5 of @xcite ) , hence it is a frchet space .
[ luno ] if @xmath108 is a sequence in @xmath60 converging to @xmath109 in @xmath60 , then there exists a subsequence @xmath110 that converges to @xmath109 almost everywhere .
if @xmath0 is compact , the claim is clear .
if not , take the increasing sequence of compact subsets @xmath1272 defining the fundamental family of semi - norms .
the topology of @xmath60 is such that @xmath136 converges to @xmath109 in @xmath1273 for all @xmath1268 .
we procede by induction on @xmath1274 .
suppose that we have found a subsequence @xmath1275 and a negligible subset @xmath1276 such that @xmath1277 clearly @xmath1275 converges to @xmath109 in @xmath1278 and we can further extract a subsequence @xmath1279 for which there exists a negligible subset @xmath1280 such that @xmath1281 set @xmath1282 and @xmath1283 . given @xmath180 , fix @xmath1284 such that @xmath1285 , so that @xmath1286 for all @xmath1287 .
then @xmath1288 is a subsequence of @xmath1289 which converges to @xmath112 . given @xmath67 , @xmath1290 since a subset @xmath1291 is negligible if and only if @xmath1292 is negligible .
notice that @xmath1293 since @xmath53 is continuous , @xmath60 is invariant under the map @xmath1294 .
let @xmath47 be a locally convex space with a saturated fundamental system @xmath130 of semi - norms and @xmath191 a ( linear ) representation of @xmath0 on @xmath47 .
a. [ sepcontrep ] the representation @xmath191 is separately continuous if a. for all @xmath66 , @xmath1295 is continuous from @xmath47 to @xmath47 ; b. for all @xmath218 , @xmath1296 is continuous from @xmath0 into @xmath47 .
b. [ contrep ] the representation @xmath191 is continuous if a. if @xmath1297 is continuous from @xmath1298 into @xmath47 . c.
the representation @xmath191 is equicontinuous if a. if @xmath1297 is continuous from @xmath1298 into @xmath47 ; b. @xmath1299 is equicontinuous , _
i.e. _ for every semi - norm @xmath1300 there exists a semi - norm @xmath1301 and a constant @xmath174 such that @xmath1302 for all @xmath66 .
if @xmath47 is a frchet space , then [ sepcontrep ] ) implies [ contrep ] ) ( proposition 1 chapter viii.2.1 of @xcite ) .
furthermore , @xmath191 is continuous if and only if for any compact set @xmath376 of @xmath0 , @xmath1303 is equicontinuous and the map @xmath1296 is continuous for all @xmath1304 , where @xmath203 is a total set in @xmath47 ( remark 2 of definition 1 chapter viii.2.1 of @xcite ) . the basic property of convolution is given by the following lemma .
if @xmath74 exists , it is a @xmath52-measurable function whose equivalence class in @xmath31 depends only on the equivalence classes of @xmath109 and @xmath125 . without loss of generality , we can suppose that both @xmath109 and @xmath125 are positive .
the topological isomorphism @xmath1305 , @xmath1306 has the property that a set @xmath1307 is @xmath1308-negligible if and only if @xmath1309 is @xmath1308-negligible .
indeed , take @xmath47 a borel measurable subset of @xmath1310 , then @xmath1311 where @xmath1312 and @xmath1313 .
hence @xmath1314 if and only if @xmath1315 for almost all @xmath66 , which is equivalent to the fact that @xmath1316 for almost all @xmath66 , _
i.e. _ @xmath1317 . as a consequence ,
the map @xmath1318 is @xmath1319-measurable , and if we change @xmath1320 on a negligible set , @xmath138 will change on a negligible set , too . since @xmath0 is second countable , the measure @xmath52 is moderated and proposition 7.b ) chapter v.8.3 of @xcite shows that the map @xmath1321 is @xmath52-measurable , where the integral is finite by assumption .
therefore , @xmath1322 depends only on the equivalence class of @xmath109 and @xmath125 .
if @xmath1323 , @xmath74 exists and @xmath1324 is in @xmath60 , then we say that @xmath109 and @xmath125 are _
convolvable_. since @xmath1323 , then @xmath1325 and @xmath1326 are ( radon ) measures on @xmath0 .
the fact that @xmath109 and @xmath125 are convolvable is equivalent to the fact that @xmath995 and @xmath1327 admit a convolution , _
i.e. _ for all @xmath1181 the function @xmath1328 is @xmath1329-integrable , namely @xmath1330 the two definitions agree , since : a. if @xmath995 and @xmath1327 admit a convolution , the map @xmath1331 defines a measure on @xmath0 whose density is precisely @xmath74 ( proposition 10 chapter viii.3.2 of @xcite and proposition 10 chapter viii.4.5 of @xcite ) ; b. if @xmath1324 exists and is in @xmath60 , then @xmath995 and @xmath1327 admit a convolution ( proposition 9 chapter viii.4.5 of @xcite ) .
we recall the following sufficient conditions .
a. corollary 20.14 of @xcite : if @xmath84 and @xmath1332 with @xmath150 $ ] , then @xmath109 and @xmath125 are convolvable , @xmath74 belongs to @xmath69 and @xmath1333 b. theorem 20.18 of @xcite : if @xmath1334 , @xmath1335 and @xmath1336 with @xmath1337 , @xmath1338 satisfying @xmath1339 with @xmath828 , then @xmath109 and @xmath125 are convolvable and @xmath1340 belongs to @xmath1341 .
furthermore , if @xmath1342 , then @xmath1343 c. theorem 20.16 of @xcite : under the same assumptions on @xmath109 and @xmath125 as in the previous item , if @xmath1344 with @xmath516 , then @xmath109 and @xmath125 are convolvable and @xmath1345 belongs to @xmath1346 and @xmath1347 d. [ young ] theorem 20.16 of @xcite : if @xmath84 and @xmath1348 ( which is equivalent to @xmath1349 ) or if @xmath1350 and @xmath1351 , then @xmath109 and @xmath125 are convolvable , @xmath1345 is a bounded continuous function , and @xmath1352 in general , the convolution is not associative .
we recall a sufficient condition as well as some other useful relations . given @xmath1353 , @xmath1354 and , for all @xmath66
, @xmath1355 provided that one of the two sides of each equality exists .
if @xmath1356 are such that either @xmath1324 and @xmath1357 exist or @xmath1358 and @xmath1359 exist , then @xmath1360 and all the convolutions exist . to prove just compute @xmath1361 next we prove .
fubini theorem gives that , for all @xmath1362 @xmath1363 by the change of variable @xmath1364 .
hence the two assumptions implies that the map @xmath1365 is in @xmath1185 . since @xmath1366 ,
all the convolutions in exist and fubini theorem implies the claimed equality .
the remaining assertions are standard .
let @xmath190 be a locally convex topological vector space , and let @xmath572 be a hausdorff locally compact second countable topological space with a positive measure @xmath755 , which is finite on all compact subsets .
a function @xmath1367 is called scalarly integrable if the scalar function @xmath1368 is integrable for every @xmath205 .
if @xmath229 is scalarly integrable , the map @xmath1369 defines a linear functional on @xmath193 , possibly not continuous ; that is , there exists an element in the algebraic dual @xmath1370 , called the scalar integral of @xmath229 and denoted @xmath1371 such that @xmath1372 usually one is interested to understand under which conditions the scalar integral lies in @xmath47 . in our paper we often look ar the case in which the argument takes its values in a dual space ( or in a space which embeds into a dual space ) , namely @xmath1373 where @xmath1374 is the space @xmath1375 endowed twith the weak*-topology , namely the topology of simple convergence , so that @xmath1376 . a locally convex space @xmath190 is said to have the property ( gdf ) if every linear map from @xmath190 to a banach space which has sequentially closed graph is actually continuous ( that is , the closed graph theorem holds true for banach space - valued linear maps defined on @xmath190 ) .
all the frchet spaces enjoy the property ( gdf ) ( @xcite , chapter i.3.3 , corollary 5 ) .
also , the dual space of any reflexive frchet space satisfies the property with respect to the strong topology , namely the topology of the convergence on bounded sets ( @xcite , chapter 6 , appendix , n@xmath1377 2 , proposition 3 ) .
the key theorem for the convergence of scalar integrals with values in a dual vector space is the following .
[ gelfand - dunford ] let @xmath190 be a hausdorff locally convex topological vector space with the property ( gdf ) .
then , for any scalarly integrable function @xmath1378 , we have @xmath1379 in this final section we recall , for the reader s convenience , the main results obtained in @xcite , specialized to our setting .
set @xmath567 and define @xmath1380 with the initial topology , which makes each inclusion @xmath1381 continuous , and @xmath1382 with the final topology , which makes each inclusion @xmath1383 continuous .
[ classic ] the space @xmath1 is a reflexive frchet space and @xmath612 is a complete reflexive locally convex topological vector space . for each @xmath1384 ,
the linear map @xmath1385 is continuous and @xmath1386 identifies , as topological vector spaces , the dual of @xmath1 with @xmath612 . for each @xmath1387 ,
the linear map @xmath1388 is continuous and @xmath1166 identifies , as topological vector spaces , the dual of @xmath612 with @xmath1 .
here we refer to @xcite .
observe that the haar measure @xmath52 is @xmath1389-finite since @xmath0 is locally compact and second countable and @xmath52 is finite on compact subsets .
furthermore , denoted by @xmath1390 , clearly @xmath1391 .
proposition 2.1 and the following remark show that the map @xmath1392 is a topological isomorphism from @xmath1 onto the strong dual of @xmath612 .
theorem 2.1 and corollary 3.2 show that the map @xmath1393 is a topological isomorphism from @xmath612 onto the strong dual of @xmath1 .
hence we can identify , as topological vector spaces , the dual of @xmath1 with @xmath612 and the dual of @xmath612 with @xmath1 .
so that both @xmath1 and @xmath612 are reflexive locally convex vector spaces .
theorem 3.1 proves that @xmath1 is a frecht space and corollary 3.3 shows that @xmath612 is complete .
note that , since @xmath1 and @xmath612 are reflexive spaces , they are barrelled ( theorem 2 iv.2.3 of @xcite ) .
s. dahlke acknowledges support from deutsche forschungsgemeinschaft ( dfg ) , grant da 360/191 . f. de mari , e. de vito and s. vigogna
were partially supported by progetto prin 2010 - 2011 `` variet reali e complesse : geometria , topologia e analisi armonica '' .
d. labate acknowledges support from nsf grants dms 1008900 and 1005799 .
s. dahlke , g. steidl and g. teschke were partially supported by daad project 57056121 ,
hochschuldialog mit sdeuropa 2013 .
h. g. feichtinger , k. h. grchenig , a unified approach to atomic decompositions via integrable group representations , in : function spaces and applications ( lund , 1986 ) , vol .
1302 of lecture notes in math . ,
springer , berlin , 1988 , pp .
5273 .
s. dahlke , g. kutyniok , p. maass , c. sagiv , h. g. stark , g. teschke , the uncertainty principle associated with the continuous shearlet transform , int . j. wavelets multiresolut .
inf . process . 6 ( 2008 ) 157181 .
s. dahlke , g. steidl , g. teschke , multivariate shearlet transform , shearlet coorbit spaces and their structural properties , in : shearlets : multiscale analysis for multivariate data , applied and numerical harmonic analysis , springer , 2012 , pp .
105144 .
e. hewitt , k. a. ross , abstract harmonic analysis .
i : structure of topological groups .
integration theory , group representations , die grundlehren der mathematischen wissenschaften , bd . 115 , academic press inc . , publishers , new york , 1963 | we set up a new general coorbit space theory for reproducing representations of a locally compact second countable group @xmath0 that are not necessarily irreducible nor integrable .
our basic assumption is that the kernel associated with the voice transform belongs to a frchet space @xmath1 of functions on @xmath0 , which generalizes the classical choice @xmath2 .
our basic example is @xmath3 , or a weighted versions of it . by means of this choice it is possible to treat , for instance , paley - wiener spaces and coorbit spaces related to shannon wavelets and _ schrdingerlets_. |
recently , the lattice boltzmann(lb ) method got substantial progress and has been regarded as a promising alternative for simulating many complex phenomena in various fields@xcite . unlike the macroscopic computational fluid dynamics or the microscopic molecular dynamics , the lb uses a mesoscopic discrete boltzmann equation to describe the fluid system . because of its intrinsic kinetic nature , the lb contains more physical connotation than navier - stokes or euler equations based on the continuum hypothesis@xcite . from the chapmann - enskog analysis , the latter can be derived from the former under the hydrodynamic limit . although having achieved great success in simulating incompressible fluids , the application of lb to high - speed compressible flows still needs substantial effort .
high - speed compressible flows are ubiquitous in various fields , such as explosion physics , aeronautics and so on@xcite .
simulation of the compressible navier - stokes system , especially for the those containing shock waves or contact discontinuities , is an interesting and challenging work . along the line
, extensive efforts have been made in the past years .
alexander , et al@xcite presented a model where the sound speed is selectable ; yan , et al@xcite proposed a compressible lb model with three - speed - three - energy - level for the euler system ; yu and zhao@xcite composed a model for compressible flows by introducing an attractive force to soften sound speed ; sunsch_pre_1998,sch_pre_2000,sch_pre_2003 contributed a locally adaptive semi - discrete lb model , where the set of particle speed is chosen according to the local fluid velocity and internal energy so that the fluid velocity is no longer limited by the particle speed set . in the development of lb for navier - stokes systems , another way is referred to as the finite difference lattice boltzmann method ( fdlbm)@xcite .
the one by watari - tsutahara ( wt ) is typical @xcite . the same idea was then extended to binary compressible flows@xcite .
fdlbmwatari_pre_2003,xag2005 breaks the binding of discretizations of space and time and makes the particle speeds more flexible . but
similar to previous lb models , the numerical stability problem remains one of the few blocks for its practical simulation to high mach number flows .
the stability problem of lb has been addressed and attempted for some years lbe-1,yong2003,xiong2002,tosi2006,ansumali2002,li2004,sofonea2004,brownlee2007,seta , pxzj , noi . among them
, the entropic lb method@xcite tries to make the scheme to follow the @xmath0-theorem ; the fix - up methodtosi2006,li2004 is based on the standard bgk scheme , uses a third order equilibrium distribution function and a self - adapting updating parameter to avoid negativeness of the mass distribution function .
flux limiter techniques are used to enhance the stability of fdlb by sofonea , et alsofonea2004 .
adding minimal dissipation locally to improve stability is also suggested by brownlee , et al@xcite , but there such an approach is not explicitly discussed .
all the above mentioned attempts are for low mach number flows . in this paper
we present a new lb scheme for high - speed compressible flows which is composed of three components , ( i ) the original dvm by wt , ( ii ) an modified lax - wendroff ( mlw ) finite difference scheme where reasonable dissipation and dispersion are naturally included , ( iii ) additional artificial viscosity . with the new scheme ,
high speed compressible flows with strong shocks can be successfully simulated .
this paper is organized as follows . in section 2 the original dvm by wt
is briefly reviewed .
an alternative fd scheme combined with artificial viscosity is introduced in section 3 .
the von neumann stability analysis is performed in section 4 , from which solutions to improve the numerical stability can be found .
several benchmark tests are used to validate the proposed scheme in section 5 .
section 6 presents the concluding remarks .
dvm of wt can be write as : @xmath1,i=1,2 ... 8\text { , } \label{dvm_eq_1}\]]where subscript @xmath2 indicates the @xmath2-th group of the particle velocities whose speed is @xmath3 and @xmath4 indicates the direction of particle s speed .
a sketch of the dvm is referred to fig.1 .
it s easy to prove that this dvm at least up to seventh rank isotropy .
the evolution of the distribution function @xmath5 with the bhatanger - gross - krook approximation@xcite reads , @xmath6 , \label{bgk_eq}\]]where @xmath7 is the discrete version of the local equilibrium distribution function ; @xmath8 is the spatial coordinate ; @xmath9 is the relaxation time ; the local particle density @xmath10 , hydrodynamic velocity @xmath11 and temperature @xmath12 are defined by @xmath13@xmath14@xmath15where @xmath16 and @xmath17are the local pressure and internal energy .
this model is designed to recover the following navier - stokes equations @xmath18@xmath19=0 , \label{ns_eq_2}\]]@xmath20+\frac{\partial } { \partial r_{\alpha } } [ \rho u_{\alpha } ( e_{int}+\frac{u^{2}}{2}+\frac{p}{% \rho } ) ] -\frac{\partial } { \partial r_{\alpha } } [ \kappa ^{^{\prime } } \frac{% \partial e_{int}}{\partial r_{\alpha } } \notag \\
+ \mu u_{\beta } ( \frac{\partial u_{\beta } } { \partial r_{\alpha } } + \frac{% \partial u_{\alpha } } { \partial r_{\beta } } -\frac{\partial u_{\gamma } } { % \partial r_{\gamma } } \delta _ { \alpha \beta } ) ] = 0 \label{ns3_eqq}\end{gathered}\]]in the hydrodynamic limit , where @xmath21 , @xmath22 are viscosity coefficient and heat conductivity coefficient , having the following relations with pressure @xmath16 and relaxation time @xmath9 : @xmath23the equilibrium distribution function @xmath7 is calculated in the following way , @xmath24 + \notag \\ & & \frac{v_{ki\varepsilon } v_{ki\pi } v_{ki\vartheta } u_{\varepsilon } u_{\pi } u_{\vartheta } } { 6\theta ^{3}}+\frac{v_{ki\varepsilon } v_{ki\pi } v_{ki\vartheta } v_{ki\xi } u_{\varepsilon } u_{\pi } u_{\vartheta } u_{\xi } } { % 24\theta ^{4 } } ] \label{feq_eq}\end{aligned}\]]with@xmath25\text { , } \label{fk_eq1}\end{aligned}\]]@xmath26and @xmath27where @xmath28we choose @xmath29 and four nonzero @xmath30 .
to simplify the discussion , we work on the general cartesian coordinate . the combination of the above dvm and the general fd scheme with first - order forward in time and second - order upwinding in space composes the original fdlb model by wt .
it has been validated via the couette flow , small mach number riemann problems .
when the mach number @xmath31 exceeds @xmath32 the original lb model is not stable .
the dvm is derived independent of mach number .
therefore , we resort to the discretization of the left - hand side of eq . to make an accurate and stable lb scheme . here
we investigate a mixed scheme which is composed of a modified lax - wendroff@xcite and an artificial viscocity . as we know , the original lax - wendroff ( lw ) scheme is very dissipative and has a strong smoothing effect " . obviously , it is not favorable when needing capture shocks in the system . to compromise the accuracy and stability ,
we add a dispersion term and the artificial viscosity to the right - hand side of eq . before discretization
so that we have @xmath33 + % \frac{v_{ki\alpha } ( 1-c_{ki\alpha } ^{2})\delta r_{\alpha } ^{2}}{6}\frac{% \partial ^{3}f_{ki}}{\partial r_{\alpha } ^{3 } } \notag \\ & & + \theta _ { \alpha i}\left\vert \kappa _ { \alpha } \right\vert ( 1-\left\vert \kappa _ { \alpha } \right\vert ) \frac{\delta r_{\alpha } ^{2}}{2\delta t}\frac{% \partial ^{2}f_{ki}}{\partial r_{\alpha } ^{2}}\text { , } \label{mbe}\end{aligned}\]]where @xmath34@xmath35plays a role of the switching function , @xmath36 is a coefficient controlling the amplitude of the artificial viscosity . using the lax - wendroff to the left - hand side and central difference to the right - hand side of eq .
results in the following lb equation , @xmath37 \notag \\ & & + \frac{c_{ki\alpha } ^{2}}{2}(f_{kii+1}-2f_{kii}+f_{kii-1 } ) \notag \\ & & + \frac{c_{ki\alpha } ( 1-c_{ki\alpha } ^{2})}{12}% ( f_{kii+2}-2f_{kii+1}+2f_{kii-1}-f_{kii-2 } ) \notag \\ & & + \frac{\theta _ { \alpha i}\left\vert \kappa _ { \alpha } \right\vert ( 1-\left\vert \kappa _ { \alpha } \right\vert ) } { 2}% ( f_{kii+1}-2f_{kii}+f_{kii-1})\text { , } \label{dis_eq1}\end{aligned}\]]where the third suffixes @xmath38 indicate the mesh nodes in @xmath39 or @xmath40 direction . the positions of terms 3 and 4 in the right - hand side of eq . have been exchanged .
it is clear that the first line corresponds to the general lb equation with the central difference in space ; compared with the central difference , the lax - wendroff contributes an extra line ii ; lines iii and iv show the added dispersion term and artificial viscosity .
we analysis the numerical stability of the fdlbm by means of von neumann stability analysis@xcite . in the analysis solution of finite - difference equation
is written as the familiar fourier series , and the numerical stability is evaluated by the magnitude of eigenvalues of an amplification matrix .
the small perturbation @xmath41 is defined as : @xmath42 , where @xmath43 is the global equilibrium distribution function which is a constant , depends only on the mean density , velocity and temperature . from equation
we can obtain @xmath44 \notag \\
+ \frac{v_{ki\alpha } ( 1-c_{ki\alpha } ^{2})\delta r_{\alpha } ^{2}}{6}\frac{% \partial ^{3}\delta f_{ki}}{\partial r_{\alpha } ^{3}}+\theta _ { \alpha i}\left\vert \kappa _ { \alpha } \right\vert ( 1-\left\vert \kappa _ { \alpha } \right\vert ) \frac{\delta r_{\alpha } ^{2}}{2\delta t}\frac{\partial ^{2}\delta f_{ki}}{\partial r_{\alpha } ^{2}}\text{. } \label{delt_fki_eq}\end{gathered}\]]the perturbation part @xmath45 may be written as series of complex exponents , @xmath46 , where @xmath47 is an amplitude at grid point @xmath48 and time @xmath49 , @xmath50 is an imaginary unit , and @xmath51 is the wave number of sine wave in the domain with the highest resolution @xmath52 . substituting this expansion into the equation , we obtain @xmath53 , where @xmath54 is a matrix being used to assess amplification rate of @xmath47 per time step @xmath55 . if the maximum of the eigenvalues of the amplification matrix satisfies the condition , @xmath56 , for all wave numbers , the fd scheme is surely stable , where @xmath57 is the eigenvalue of the amplification matrix .
this is the von neumann condition for stability .
the amplification matrix @xmath54 can be written as following,@xmath58@xmath59several researchers have analyzed the stability of the incompressible lb models@xcite , it is found that there is not a single wave - number being always the most unstable .
for the 2d dvm by wt , @xmath60 is a matrix with @xmath61 elements .
moreover , every element is related to the macroscopical variables ( density , temperature , velocities ) , discrete velocities and other constants , so it is difficult to analyze with explicit expressions .
we resort to using the software , mathematica-5 to conduct a series of quantitative analysis .
now we show some numerical results of von neumann analysis by mathematica-5 .
the results will be shown by figures with curves for the maximum eigenvalue @xmath62 of @xmath60 versus @xmath63 .
figure 2 shows a comparison of the fours different cases , ( i ) lb with only the central - difference to the convection term , i.e. lb scheme based only on the first line of eq .
( see the black line with squares ) ; ( ii ) lb with lax - wendroff , i.e. , lb scheme based on the first two lines of eq .
( see the red line with circles ) ; ( iii ) lb scheme based on lines 1 and 3 of eq .
( see the green line with triangles ) ; ( iv ) lb scheme based on the whole of eq .
( see the blue line with triangles down ) . here
the macroscopic variables are chosen as @xmath64 = @xmath65 , and the remaining parameters are @xmath66 = @xmath67 , @xmath68 , @xmath69 , @xmath70 .
it is clear that the artificial viscosity term can significantly decrease the maximum eigenvalue @xmath71 from being larger than to be smaller than @xmath72 for appropriately given time step .
moreover , it is worthy to note that dissipation term in line 2 of eq .
favors and dispersion term in line 3 disfavors the stability to some extent .
numerical experiments show that the dispersion term may effectively reduce numerical oscillations near discontinuity and improves the accuracy ( see fig . 3 for an example ) .
= @xmath73 , the other constants are set as @xmath66 = @xmath67 , @xmath74,@xmath69,@xmath75 @xmath76 @xmath77.,scaledwidth=67.0% ] , @xmath78 , @xmath79 .
the other constants and macroscopic variables are the same as in fig.2.,scaledwidth=67.0% ] figure 4 shows the effects of various artificial viscosities to the stability .
fig.(a ) shows the cases with @xmath80 , @xmath81 , @xmath82 , and @xmath83 .
fig.(b ) shows the cases with @xmath80 , @xmath84,@xmath85 , and @xmath86 .
the other constants and macroscopic variables are unchanged . from this figure
we can see some relevance : strength of artificial viscosity has a large impact on the stability .
lb works only within a certain range of artificial viscosity . in practical simulations ,
we generally take the smaller viscosity in favor of the accuracy . ,
@xmath87 , @xmath82 , @xmath88 .
( b ) shows the cases with @xmath89,@xmath90 , @xmath85 , @xmath86 .
the other constants and macroscopic variables are unchanged.,scaledwidth=90.0% ] since the density @xmath91 can be normalized to @xmath72 , we then need only investigate the effects of the other two physical quantities , temperature @xmath12 and flow velocity @xmath92 .
figure 5 shows four cases with @xmath93 , @xmath94 , @xmath95 and @xmath96 . here
@xmath97 , @xmath98 and @xmath99 .
when other parameters are fixed , the numerical stability becomes better with the increasing of temperature .
this can also be understood that higher temperature corresponds to higher sound speed and lower mach number . to numerical stability .
@xmath97 , @xmath98 and @xmath100 .
the other physical quantities and model parameters are unchanged.,scaledwidth=67.0% ] to numerical stability .
the value of @xmath101 is altered from zero to @xmath102 and @xmath98 .
here @xmath100 , the other constants and macroscopic variables are unchanged.,scaledwidth=67.0% ] figure 6 shows cases with difference flow velocities .
the value of @xmath101 is altered from zero to @xmath102 and @xmath98 .
here @xmath99 , the other constants and macroscopic variables are unchanged .
this figure clearly shows that the higher the mach number , the larger the maximum eigenvalue , which answers why the numerical stability becomes worse with the increasing of mach number of the fluid .
in this section two kinds of typical benchmarks are used to validate the newly proposed scheme .
the first kind is the riemann problem@xcite .
the second one is the problem of shock reflection@xcite . here
the two - dimensional model is used to solve the one - dimensional riemann problem .
the initial macroscopic variables at the two sides are denoted by @xmath103 , @xmath101 , @xmath104 , @xmath105 and @xmath106 , @xmath101 , @xmath104 , @xmath107 , respectively . for the problem considered ,
the initial condition is described by @xmath108figure 7 shows the computed density , pressure , velocity , temperature profiles at @xmath109 , where the circles are simulation results and solid lines with squares are analytical solutions .
the size of grid is @xmath110 , time step @xmath111 , and @xmath112 @xmath113 , @xmath114 the two sets of results have a satisfying agreement .
the initial condition of this problem reads @xmath115figure 8 shows the results at @xmath109 , where the circles are simulation results and solid lines with squares correspond to exact solutions .
the parameters are set to be @xmath116 , @xmath111 , @xmath117
@xmath113 , and @xmath118 . we also find a good agreement between the two sets of results . .
solid lines with squares are for exact solution and circles are for simulation results.,scaledwidth=80.0% ] the initial condition of this problem is @xmath119the numerical and exact solutions at @xmath120 are shown in fig.9 , where @xmath121 , @xmath70 and @xmath122 .
the exact solution of this problem consists of two strong rarefaction waves and a weak constant contact discontinuity .
pressure near the contact discontinuity is very small , which brings certain difficulties to simulation .
temperature and density calculated by many schemes are negative .
however , the improved model ensures the positivity of them .
successful simulation of this problem proves that the improved model is applicable to the low - density , low - temperature flow simulations . .
solid lines with squares are for exact solutions , and circles are for simulation results.,scaledwidth=80.0% ] the initial condition of this test can be write as @xmath123this is generally regarded as a difficult test .
the exact solution contains a leftwards rarefaction wave , a contact discontinuity and a strong shock .
it is generally used to check the robustness and accuracy .
figure 10 gives comparison of the numerical and theoretical results at @xmath124 .
here @xmath125 , other parameters are same as in the sjogreen test .
successful simulation of this test proves the improved model is applicable to flows with very high ratios of temperature and pressure .
this test with the following initial data : @xmath126this is also a difficult test .
exact solution contains a leftwards shock , a right contact discontinuity and shock which spreading to right side . and the left - shock spreads to right very slowly , which brings additional difficulties to the numerical method .
fig.11 gives a comparison of the numerical and theoretical results at @xmath127 .
parameters used in this test are @xmath128 , @xmath70 and @xmath129 .
the good agreement between the two sets of results shows again the robustness of the improved model . .
solid lines with symbols are for exact solutions and symbols are for simulation results.,scaledwidth=80.0% ] we will present two gas dynamics simulations . both are done on rectangular grid .
the first is to recover a steady regular shock reflection .
the second is the double mach reflection of a shock off an oblique surface .
this example is used in ref.@xcite as a benchmark test for comparing the performance of various difference methods on problem involving strong shocks . in the first test problem , the incoming shock wave with mach number 20
has an angle of @xmath130 to the wall .
the computational domain is a rectangle with length @xmath131 and height @xmath132 .
this domain is divided into a @xmath133 rectangular grid with @xmath134 .
the boundary conditions are composed of a reflecting surface along the bottom boundary , supersonic outflow along the right boundary , and dirichlet conditions on the left and top boundary conditions , given by @xmath135figure 12 shows contours of density at @xmath109 .
the clear shock reflection on the wall agrees well with the exact solution . in this test
, we considered an unsteady shock reflection .
the initial pressure ratio here is high .
a planar shock is incident towards an oblique surface with a @xmath130 angle to the direction of propagation of the shock .
a uniform mesh size of @xmath136 is used for the numerical simulation .
the conditions for both sides are : ( bottom ) of the double mach reflection problem at the 750th iteration step .
the units of the @xmath39- and @xmath40- axes are both @xmath137 .
, title="fig:",scaledwidth=80.0% ] ( bottom ) of the double mach reflection problem at the 750th iteration step .
the units of the @xmath39- and @xmath40- axes are both @xmath137 .
, title="fig:",scaledwidth=80.0% ] [ width=0.8]fig13c.eps @xmath138 where @xmath139 .
the reflecting wall lines along the bottom of the problem domain , beginning at @xmath140 the shock makes a @xmath141 angle with the @xmath39 axis and extends to the top of the problem domain at @xmath142 at the top boundary , the physical quantities are assigned the same values as on the left side for @xmath143 and are assigned the same values on the right side , where @xmath144 .
the computed density , temperature and flow velocity along the @xmath39-direction are shown in fig.13 , where complex characteristics , such as oblique shocks and triple points , are well captured .
a lattice boltzmann model to the high - speed compressible navier - stokes system is presented .
the new lb model is composed of the following components : the original dvm by watari and tsutahara , a modified lax - wendroff scheme and an additional artificial viscosity . compared with the central difference scheme
, the lax - wendroff contributes a dissipation term which is in favor of the numerical stability , even though it is generally still not enough for high - speed flows .
the introducing of the third - order dispersion term helps to eliminate some unphysical oscillations at discontinuity .
the additional artificial viscosity compensates the insufficiency of the above - mentioned dissipation so that the lb simulation can continue smoothly .
the adding of the dispersion and artificial viscosity terms should survive the dilemma of stability versus accuracy .
in other words , they should be minimal but make the evolution satisfy the von neumann stability condition . due to the complexity
, the analysis resorts to the software , mathematica-5 , and only some typical results are shown by figures .
typical benchmark tests are used to validate the proposed scheme .
riemann problems , including the sod , lax , sjogreen , colella explosion wave , collision of two strong shocks , show good accuracy and numerical stability of the new scheme , even though they are generally difficult to resolve by traditional computational fluid dynamics@xcite . regular and double
mach shock reflections are successfully recovered .
these simulations show that the improved lb model may be used to investigate some long - standing problems , such as the transitions between regular and mach reflections . by incorporating an appropriate equation of state , or equivalently , a free energy functional , or an external force
, the present model may be used to simulate liquid - vapor transition and relevant flow behavior .
future work includes a more complete description of the problem on numerical accuracy versus stability and thermal lattice boltzmann model for multi - phase flows .
this work is supported by the national basic research program ( 973 program ) [ under grant no .
2007cb815105 ] , science foundation of laboratory of computational physics , national natural science foundation [ under grant nos .
10775018,10474137 and 10702010 ] of china .
r. benzi , s. succi , and m. vergassola , phys . rep . * 222 * , 145 ( 1992 ) ; d. a. wolf - gladrow , _ lattice gas cellular automata and lattice boltzmann models _ , springer - verlag , new york ( 2000 ) ; h. chen , s. kandasamy , s. orszag , r. shock , s. succi , and v. yakhot , science 301 , 633 ( 2003 ) .
aiguo xu , g. gonnella , and a. lamura , phys .
e * 74 * 011505(2006 ) ; phys .
e * 67 * , 056105 ( 2003 ) ; physica a * 331 * , 10 ( 2004 ) ; physica a * 344 , * 750 ( 2004 ) ; physica a * 362 * , 42 ( 2006 ) ; aiguo xu , commun .
. phys . * 39 * , 729 ( 2003 ) . | in this paper we present an improved lattice boltzmann model for compressible navier - stokes system with high mach number .
the model is composed of three components : ( i ) the discrete - velocity - model by watari and tsutahara [ phys rev e * 67*,036306(2003 ) ] , ( ii ) a modified lax - wendroff finite difference scheme where reasonable dissipation and dispersion are naturally included , ( iii ) artificial viscosity .
the improved model is convenient to compromise the high accuracy and stability .
the included dispersion term can effectively reduce the numerical oscillation at discontinuity .
the added artificial viscosity helps the scheme to satisfy the von neumann stability condition .
shock tubes and shock reflections are used to validate the new scheme . in our numerical tests
the mach numbers are successfully increased up to 20 or higher .
the flexibility of the new model makes it suitable for tracking shock waves with high accuracy and for investigating nonlinear nonequilibrium complex systems .
lattice boltzmann method , compressible flows , von neumann analysis 47.11.-j , 51.10.+y , 05.20.dd |
two - dimensional honeycomb lattice of carbon atoms may be viewed as the mother of all other forms of carbon .
its crucial electronic property , which arises as a consequence of the absence of the inversion symmetry around the lattice site , is that the usual fermi surface is reduced to just two points .
the electronic dispersion may be linearized around these two points , after which it becomes isotropic and dependent on the single dimensionful parameter , fermi velocity @xmath4 .
the pseudo - relativistic nature of the electronic motion in graphene has since its synthesis placed this material at the center stage of condensed matter physics .
many qualitatively novel phenomena that take , or may take place in such a system of `` dirac '' electrons are actively discussed in the rapidly growing literature on the subject .
@xcite in this paper we discuss the low - energy theory and the metal - insulator quantum phase transitions of the _ interacting _ dirac electrons on the honeycomb lattice , building upon and expanding significantly the earlier work by one of us .
@xcite in the first approximation , _ all _ weak interactions of dirac electrons in graphene may be neglected at half filling , when the fermi surface consists of the dirac points .
this is because short - range interactions are represented by local terms which are quartic in the electron fields , which makes them irrelevant near the non - interacting fixed point by power counting .
the same conclusion turns out to apply to the long - range tail of the coulomb interaction , which remains unscreened in graphene , although only marginally so .
@xcite nevertheless , if strong enough , the same interactions would turn graphene into a gapped mott insulator . as an example , at a strong on - site repulsion the system is likely to be the usual nel antiferromagnet .
@xcite it is not _ a priori _ clear on which side of this metal - insulator transition graphene should be . with the standard estimate for the nearest - neighbor hopping in graphene of @xmath5 and the hubbard interaction of @xmath6 , it seems that the system is below yet not too far from the critical point estimated to be at @xmath7 .
@xcite if sufficiently weak , the electron - electron interactions only provide corrections to scaling of various quantities , which ultimately vanish at low temperatures or frequencies . at , what is probably a more realistic , an intermediate strength , the flow of interactions and the concomitant low - energy behavior may be influenced by the existence of metal - insulator critical points .
it is possible that some of the consequences of such interaction - dominated physics have already been observed in the quantization of the hall conductance at filling factors zero and one .
@xcite as we argued elsewhere , the anomalously large value of the minimal conductivity in graphene @xcite may be yet another consequence of the coulomb repulsion between electrons .
@xcite the above discussion raises some basic questions .
what is the minimal description of interacting electrons in graphene at `` low '' energies ?
what is the symmetry of the continuum interacting theory , and how does it constrain the number of coupling constants ?
what kinds of order may be expected at strong coupling , and what is the nature of the metal - insulator quantum phase transition ? in this paper we address these and related issues . in the rest of the introduction
we give a preview of our main results . the simplest prototypical system that exhibits
the physics of interacting dirac fermions which we seek to understand is the collection of spinless electrons interacting via short - range interactions , at half - filling . for present purposes an interaction
may be considered as `` short - ranged '' if its fourier transform at the vanishing wavevector is finite .
@xcite the least irrelevant quartic terms one can add to the non - interacting dirac lagrangian will then be local in space - time , and of course quartic in terms of the four - component dirac fields that describe the electronic modes near the two inequivalent dirac points at wavevectors @xmath8 at the edges of the brillouin zone .
the most general local quartic term in the lagrangian would be of the form @xmath9 where @xmath10 and @xmath11 are four - dimensional hermitian matrices .
the symmetry alone , however , immediately drastically reduces the number of independent couplings from the apparent @xmath12 to just fifteen .
although the point group of the honeycomb lattice is @xmath13 , the exact spatial discrete symmetry of the lagrangian is only the _ dihedral group _
@xmath1 , or the _ vierergruppe _ , which consists of the reflections through the two coordinate axis shown in fig . 1 , and the inversion through the origin .
such a small symmetry results from the very choice of two inequivalent dirac points out of six corners of the brillouin zone , which reduces the symmetry to the simple exchange of the two sublattices ( reflection around a axis ) , the exchange of dirac points ( reflection around b axis ) , and their product ( the inversion through the origin ) .
@xmath1 , the time - reversal , and the translational invariance are shown to leave fifteen possible different local quartic terms in the lagrangian . fortunately , not all of these still numerous quartic terms are independent , and there are linear constraints between them implied by the algebraic fierz identities .
@xcite the fierz transformations are rewritings of a given quartic term in terms of others , and we provide the general formalism for determining the number and the type of independent quartic couplings of a given symmetry . for the case at hand we find that spinless electrons interacting with short - range interactions on honeycomb lattice are in fact described by only _
six _ independent local quartic terms .
the inclusion of electron spin would double this number to twelve .
the linearized noninteracting lagrangian for dirac electrons , @xmath14 as well - known , exhibits the lorentz and the global @xmath3 ( `` chiral '' ) symmetry .
the latter , generated by @xmath15 , with @xmath16 , is nothing but the `` rotation '' of the `` pseudospin '' , or `` valley '' , corresponding to two inequivalent dirac points .
@xcite a general quartic term allowed by the lattice symmetry , on the other hand , has a much smaller symmetry , as already mentioned .
nevertheless , we will argue that near the metal - insulator quantum critical points , all , or nearly all of the larger symmetry possessed by the non - interacting part of the lagrangian gets restored .
this conclusion is supported by the , admittedly uncontrolled , but nevertheless quite informative one - loop calculation .
first , we find three distinct critical points in the theory , all of which have not only the rotational , but the full lorentz - symmetric form .
this immediately implies that the dynamical critical exponent is always @xmath17 .
this is quite remarkable in light of the fact that the microscopic theory is not even rotationally invariant , and that the critical points in question are purely short - ranged .
@xcite the fact that @xmath17 has important implications for several key physical observables near the critical point , as we discuss shortly .
furthermore , we find that two out of three critical points in the theory exhibit a full chiral symmetry as well .
we identify the three fixed points in the theory as corresponding to three possible order parameters , or `` masses '' that develop in the insulating phase at strong coupling .
\1 ) @xmath18 , which preserves chiral , but breaks time - reversal symmetry .
microscopically , this order parameter may be understood as a specific pattern of circulating currents , as discussed in the past .
@xcite \2 ) @xmath19 , which preserves the time - reversal symmetry , and the single chiral generator @xmath20 , which will be shown to correspond to translational invariance .
this order parameter describes a finite staggered density , i. e. the difference between the average densities on the two sublattices of the honeycomb lattice . @xcite
\3 ) @xmath21 , which preserves the time - reversal , but breaks translational invariance ( @xmath20 ) .
this order parameter can be understood as the specific `` kekule '' modulation of the nearest - neighbor hopping integrals .
@xcite in one - loop calculation all three critical points have the same correlation length exponent @xmath22 , which we believe is an artifact of the quadratic approximation .
the result that the dynamical critical exponent @xmath17 is , on the other hand , possibly exact .
if we denote the relevant interaction parameter with @xmath23 , the fermi velocity near the transition scales as @xmath24 so the above value of @xmath25 would simply imply that it stays regular at the transition .
this appears to be in agreement with the picture of the transition as the opening of the relativistic `` mass '' in the spectrum .
the mass - gap in the insulating phase scales as usual @xcite as @xmath26 the transition on the metallic side is manifested as vanishing of the residue of the quasiparticle pole @xcite @xmath27 where we assumed @xmath17 .
( a more general power - law is discussed in the text . ) at one - loop the fermion anomalous dimension @xmath28 vanishes , but in general it is a positive , small , and critical - point - dependent number . the overall picture of the metal - insulator transition that emerges is presented in fig .
2 . for graphene
s @xmath29-orbitals well localized on carbon sites , a further significant simplification takes place .
all the terms without the equal number of creation and annihilation operators for each of the two sublattices must vanish .
assuming again the emergent lorentz symmetry at low energies this allows one to finally write the simplest internally consistent interacting theory as @xmath30 this lagrangian provides the minimal low - energy description of interacting spinless electrons on honeycomb lattice .
it has two critical points , corresponding to transitions into insulators 1 ) and 2 ) in the above , both corresponding to the gross - neveu criticality in 2 + 1 dimensions .
we discuss the internal consistency and the sufficiency of this lagrangian and some of the peculiarities of the ensuing phase diagram .
the rest of the paper is organized as follows .
we discuss the point symmetry , translational symmetry , and the time - reversal symmetry of the interacting lagrangian as dictated by the microscopic hamiltonian for the system in the next section . in section iii
it is shown how further enlargements of the symmetry would reduce the number of coupling constants .
we introduce the notion of `` maximally symmetric '' theory , which shares the full lorentz and chiral symmetry with the quadratic term in the lagrangian . the general formalism of fierz transformations is developed and applied to the cases of interest in section iv .
the change of the coupling constants with the ultraviolet cutoff in the theory is studied in section v. the atomic limit of the general interacting theory is described in section vi , and the critical exponents are discussed in section vii . in section viii
we discuss the scaling of the electron propagator and the power - laws for various quantities of interest .
the discussion of the long - range coulomb interaction and the critical behavior of the specific heat and the optical conductivity are given in section ix , and the summary in section x. finally , in appendixes we present some of the requisite technical details : the fierz transformation , the spectral form of the asymmetric matrix needed in section iv , and an alternative implementation of the renormalization group in presence of linear constraints .
as the simplest microscopic model that contains the relevant physics we may consider the tight - binding hamiltonian on the graphene s honeycomb lattice , defined as @xmath31 where @xmath32 and @xmath33 are the electron annihilation operators at two triangular sublattices of the honeycomb lattice . here , @xmath34 denotes sites of the sublattice generated by linear combinations of basis vectors @xmath35 , @xmath36 , whereas @xmath37 are the sites on the second sublattice , with @xmath38 being @xmath39 , or @xmath40 , and @xmath41 is the lattice spacing . within the framework of the tight - binding model the energy spectrum
is doubly degenerate @xmath42|$ ] , and becomes linear and isotropic in the vicinity of six dirac points , at the edge of the brillouin zone , among which only two , hereafter chosen to be at @xmath43 with @xmath44 , are inequivalent . retaining only the fourier components in the vicinity of these two inequivalent points
, the quantum mechanical action corresponding to @xmath45 at low energies can be written in the form @xmath46 , with the free lagrangian density @xmath47 defined as in eq .
( 2 ) , with @xmath48 as the imaginary time and @xmath49 is the temperature .
matrices @xmath50 satisfy the clifford algebra @xmath51 , @xmath52 , @xmath53 .
the summation over repeated space - time indices is assumed hereafter .
the fermionic field @xmath54 is defined as @xmath55 here , the reference frame is conveniently rotated so that @xmath56 , @xmath57 , @xmath58 are the fermionic matsubara frequencies , @xmath59 is a high - energy cutoff , and we set @xmath60 , where @xmath61 is the fermi velocity . choosing @xmath62 implies @xmath63 the two remaining anticommuting matrices may then be taken as @xmath64 this defines the `` graphene representation '' of @xmath65-matrices .
@xcite @xmath66 are the standard pauli matrices . considering a more general model with further hoppings or weak anisotropies @xcite
can be seen not to destroy the dirac points , but only to shift them in energy . as this can always be compensated by a shift of the chemical potential
, the lagrangian ( 2 ) provides the low - energy description of the general free electronic hamiltonian on a honeycomb lattice , with the chemical potential tuned to the dirac point .
note that the free lagrangian besides the lorentz symmetrt also possesses another chiral , pseudospin or `` valley '' , global @xmath3 symmetry , generated by @xmath15 .
both the lorentz and the chiral symmetry of the free lagrangian are emerging only at low energies , however , and the term quadratic in derivatives in @xmath47 , for example , would spoil it .
as will be shown shortly , both symmetries are also violated by the leading irrelevant quartic couplings introduced by the interactions .
let us now consider the electron - electron interactions .
the hamiltonian of a general four - fermion interaction has the form @xmath67 where @xmath68 or @xmath33 are fermionic annihilation operators , and the matrix element corresponding to the interaction potential @xmath69 is given by @xmath70 here we can take @xmath71 to be a localized @xmath29-orbital on the site @xmath72 , so that it belongs to either one of the two sublattices of the honeycomb lattice . in general , there is no restriction on the overlap of the wave - functions , and all the matrix elements @xmath73 are in principle finite .
their relative sizes , however , may be rather different , and we discuss important simplifications that follow in the limit of well localized orbitals in sec .
vi . in the following
we will consider general `` short - ranged '' interactions , which are defined by the interaction @xmath74 with a regular fourier component @xmath75 . without a loss in generality one
may then take the interacting lagrangian for spinless fermions corresponding to @xmath76 as in eq .
( 1 ) where @xmath10 and @xmath11 are some constant @xmath77 hermitian matrices .
the interacting lagrangian contains therefore at most @xmath78 independent real coupling constants .
however , the number of couplings in @xmath79 is severely reduced by the lattice symmetries , as we discuss next .
two obvious discreet symmetries of the honeycomb lattice that have not been broken by our choice of the dirac points are the reflection symmetries through the lines a and b in fig .
1 . let us consider the former symmetry first .
it exchanges the two sublattices , but not the two dirac points .
the low - energy lagrangian thus has to be invariant under the exchange of the spinor components belonging to different sublattices , @xmath80 .
consequently , the symmetry operator acting on the four - component dirac spinor defined in eq .
( [ dirac - spinor ] ) has the form @xmath81 since under this reflection @xmath82 and @xmath83 , @xmath47 is evidently invariant under @xmath84 .
the invariance of @xmath79 under this reflection symmetry requires both matrices @xmath10 and @xmath11 to either commute or anticommute with the operator @xmath84 : @xmath85=[s , m_2]=0,\ ] ] or @xmath86 similarly , the reflection symmetry through the line @xmath87 exchanges the two dirac points , while not exchanging the sublattice labels .
it corresponds therefore to @xmath88 recalling that under this transformation @xmath89 and @xmath90 , it is evident that the free lagrangian in graphene representation is indeed invariant under @xmath49 as well . demanding the interacting lagrangian @xmath79 to be invariant under the action of the operator @xmath49 on the dirac spinor
, implies that both matrices @xmath10 and @xmath11 either commute or anticommute with @xmath49 as well . in other words , both matrices @xmath10 and @xmath11 have to be either even or odd with respect to @xmath49 : @xmath91=[t , m_2]=0,\ ] ] or @xmath92 together with the combination of the two reflections s and t , and the identity operation , the two symmetry operations form the _ dihedral group _ ( or klein s _ vierergrouppe _ , in older literature ) @xmath1 : @xmath93 , the symmetry group of a rectangle . note that the transformation @xmath94 is just the space inversion , and that the rotation by @xmath95 _ does not _
belong to @xmath1 .
one may now classify all the four - dimensional matrices into four categories , according to their transformation under the two reflection symmetries @xmath84 and @xmath49 , respectively : even - even , @xmath96 , even - odd , @xmath97 , odd - even , @xmath98 , and odd - odd , @xmath99 . the interacting lagrangian symmetric under the @xmath1 is thus restricted to be of the following form @xmath100 where @xmath101 , @xmath102 , @xmath103 , are real and symmetric .
the maximal number of independent real parameters specifying the allowed couplings is thus already reduced to forty , since each @xmath101 has ten independent components . the generator @xmath104 of the chiral symmetry plays a special role .
it is in fact the generator of translations . to see this recall that under a translation by @xmath105 the electron fields transform as @xmath106 where @xmath107 .
the dirac field under the same transformation thus changes as @xmath108 or , in real space , @xmath109 translational invariance requires therefore that @xmath79 is a scalar under the transformations generated by @xmath20 , which we will denote as @xmath110 .
it is easy to see that this is precisely the same as demanding the conservation of momentum in the interaction terms .
the reader is also invited to convince herself that the terms with the higher - order derivatives in @xmath47 would also be invariant under the @xmath110 .
first , we observe that there are eight linearly independent bilinears that are _ scalars _ under the @xmath110 : @xmath111 where @xmath112 and @xmath113 .
the remaining eight bilinears can be grouped into four _ vectors _ under the same @xmath110 : @xmath114 @xmath115 @xmath116 @xmath117 the invariance under @xmath110 implies therefore that the interacting lagrangian has the following form @xmath118 the number of possible independent couplings is down to eighteen .
the set of the allowed couplings is further reduced by the time - reversal symmetry . the microscopic interacting hamiltonian ( [ hint ] ) is invariant under the time - reversal , and therefore the corresponding low - energy lagrangian has to possess the same invariance .
the time - reversal symmetry requires that @xmath119 , where @xmath120 is the antiunitary operator representing the time - reversal symmetry , and thus has the form @xmath121 , with @xmath122 representing the unitary part of @xmath120 and @xmath123 is the complex conjugation . to find the form of @xmath120
let us consider first the _ massive _ dirac hamiltonian @xmath124 with the mass @xmath125 describing the imbalance in the chemical potential on the two sublattices .
@xcite recalling that momentum changes sign under the time - reversal , @xmath126 , in the graphene representation the invariance of the above hamiltonian under the same transformation implies @xmath127=[u,\gamma_0]=0,\end{aligned}\ ] ] and hence @xmath128 . within the simplest framework of the tight - binding model with uniform hopping the time - reversal operator
is not uniquely determined .
we thus consider a generalized tight - binding model with anisotropic hopping defined as @xmath129 where @xmath130 represents a non - uniform hopping , and @xmath131 .
@xcite on a lattice , this particular set of hoppings generates the so - called kekule texture .
near the two dirac points the hamiltonian @xmath132 reads @xmath133 where @xmath134 and @xmath135 .
the two masses @xmath136 and @xmath137 therefore provide the low - energy representation of a completely real microscopic hamiltonian , so that we postulate that @xmath132 is also time - reversal symmetric . in graphene representation
this requires the unitary part of the time - reversal operator to obey the following algebra : @xmath138=\{u , i{\gamma_0}{\gamma_5}\}=0.\ ] ] the matrix @xmath139 satisfies conditions ( [ condt ] ) and thus the unitary part of the operator @xmath120 acting on the spinless dirac field ( [ dirac - spinor ] ) is @xcite @xmath140 with @xmath141 as the @xmath142 unity matrix .
the unitary part of the time - reversal operator thus simply exchanges the components of the dirac spinor @xmath54 with different valley indices , as expected .
it also happens to be the same as one of the two matrices representing the reflection operators from @xmath1 .
another way of arriving at the same form for the time - reversal operator is to postulate that an arbitrary chiral transformation of the dirac hamiltonian in eq .
( 29 ) yields a time - reversal invariant hamiltonian .
alternatively , our derivation may be understood as a demonstration of commutativity of the chiral and time - reversal transformations . since we have already used the invariance under @xmath49 to restrict the interacting lagrangian ,
time - reversal invariance will be observed if the remaining terms are even under complex conjugation .
all the terms @xmath143 and @xmath144 are thus automatically invariant under time - reversal , but among the remaining six mixed terms , the terms @xmath145 , @xmath146 , and @xmath147 are odd .
time - reversal invariance implies therefore that @xmath148 which leaves then at most fifteen independent couplings .
we found that the exact symmetries of the microscopic hamiltonian , @xmath1 , translational , and the time - reversal , leave at most fifteen independent short - range couplings . anticipating some of the results ,
it is interesting to deduce the further reductions of the number of couplings if one by hand imposes larger symmetries onto the interaction lagrangian @xmath79 . since the rotation by @xmath95 is not a member of the @xmath1 , the matrices @xmath149 and @xmath150 appear asymmetrically in @xmath79 .
if we demand that they appear symmetrically , @xmath79 becomes fully rotationally invariant .
this is achieved if @xmath151 let us call the interacting lagrangian with the rotational invariance imposed this way @xmath152 .
it would be described by at most nine couplings . imposing
further the lorentz invariance would require that on top of the above restrictions one also has that @xmath153 with the restrictions in the previous two equations the lagrangian has only six couplings constants , and may be cast in a manifestly lorentz invariant form , worth displaying : @xmath154 + g_\gamma [ ( \bar{\psi } \gamma_\mu \gamma_3 \psi)^2 + ( \bar{\psi } \gamma_\mu \gamma_5 \psi)^2 ] .
\end{aligned}\ ] ] finally , the maximally invariant interacting lagrangian would be with the full , i. e. both the lorentz and the chiral , symmetry of the non - interacting part .
this is achieved by setting in the last equation @xmath155 the interacting lagrangian can in this case be written as @xmath156 where the participating bilinears in dirac fields , @xmath157 @xmath158 @xmath159 @xmath160 are the scalar ( vector ) , scalar ( scalar ) , vector ( scalar ) , and vector ( vector ) under the chiral ( lorentz ) transformation .
the last form makes the lorentz and the chiral symmetry of the lagrangian @xmath161 manifest .
such a _ maximally symmetric _
lagrangian contains therefore at most only four coupling constants .
the number of independent couplings is further reduced by the existence of algebraic identities between seemingly different quartic terms .
the derivation of the so - called fierz transformation , which allows one to write a given local quartic term in terms of other quartic terms is provided in appendix a. a systematic application of this transformation allows one to reduce the number of independent couplings for a given symmetry of the interacting lagrangian .
the application of fierz identity to the set of quartic terms allowed by the assumed symmetry in principle leads to the set of linear constraints of the form @xmath162 where @xmath163 is a real typically asymmetric matrix , and @xmath164 is a column ; the elements of which are the quartic terms allowed by the symmetry .
of course , only the quartic terms which share the same symmetry may be related by fierz transformations .
for example , in the maximally symmetric case @xmath165 .
when the number of couplings is small it is easy to discern the linearly independent combinations of the original terms , but when it is not , as the case is for @xmath166 microscopic symmetry of @xmath79 , one needs a more general method of doing so . in appendix
b we show that an asymmetric matrix such as the fierz matrix @xmath163 can be written in the diadic form @xcite as @xmath167 where @xmath168 is the real spectrum of the related symmetric matrix @xmath169 . in the eigenbasis of
@xmath170 we can write , in dirac notation , @xmath171 so that the above linear equations can be written as @xmath172 since the vectors @xmath173 also form a basis , it must be that either : \a ) for @xmath174 , @xmath175 , or \b ) @xmath176 , so that @xmath177 .
the first set provides us then with the linearly independent constraints , and the second with the set of remaining linearly independent quartic terms . since the matrices @xmath163 and @xmath170 obviously have the same kernels , the number of independent coupling constants allowed by the symmetry is simply the dimension of the kernel of the appropriate fierz matrix
let us consider the simplest example of the quartic term with the full lorentz and chiral symmetry , @xmath178 first .
defining the vector @xmath164 as in the above leads to the fierz matrix @xmath179 with the two - dimensional kernel with the zero - eigenstates @xmath180 @xmath181 and @xmath182 .
the remaining two eigenvalues are @xmath183 , and @xmath184 .
the general method explained above implies that the general maximally symmetric interacting lagrangian can be written as @xmath185 with the `` physical '' couplings @xmath186 @xmath187 the two remaining linearly independent combinations vanish due to fierz identity .
so the maximally symmetric interacting theory is specified by only two quartic coupling constants , which may be chosen to be any linearly independent combinations of the above @xmath188 and @xmath189 .
we may then proceed to find the independent couplings for the next case in order in complexity , @xmath190 , with the chiral symmetry broken down to @xmath110 .
if we define the six - dimensional vector @xmath191 the fierz matrix is found to be @xmath192 with the three - dimensional kernel spanned by @xmath193 @xmath194 @xmath195 @xmath196 can now be written in terms of only three linearly independent quartic terms , in complete analogy with the maximally symmetric case .
the symmetry ladder may now be climbed back to the @xmath166 minimally symmetric lagrangian .
first , reducing lorentz to rotational symmetry increases the number of independent couplings to four . removing the rotational symmetry finally increases the number of couplings to six . we may note in passing that there are two independent fierz identities between the mixed terms that obey the time - reversal symmetry in eq .
( 28 ) : @xmath197 @xmath198 so that the three mixed terms in fact contribute a single independent coupling .
having determined the independent coupling constants for each symmetry , we now proceed to study their changes with the decrease of the upper cutoff @xmath199
. we will be particularly interested in fixed points of such renormalization group transformations , as they will provide the information on the quantum metal - insulator transitions that can be induced by increase in interactions .
let us again begin with the maximally symmetric lagrangian , @xmath200 .
there are only two coupling constants to consider in this case , and we choose them to be @xmath201 and @xmath202 , which correspond to @xmath203 and @xmath204 quartic terms , respectively . if any of the other two terms would become generated by the renormalization transformation we would use the fierz identity to rewrite it in terms of @xmath203 and @xmath204 . alternatively , one may wish to renormalize the theory as written in terms of physical couplings in eq .
this procedure , completely equivalent to what is pursued here , is described in appendix c. as we integrate the fermionic modes lying in the 2 + 1 dimensional momentum shell @xcite from @xmath205 to @xmath199 , with @xmath206 , to quadratic order in coupling constants we find @xmath207 @xmath208 we rescaled the couplings here as @xmath209 . to this order
no other types of quartic terms actually get generated , and the fierz transformation turns out not to be necessary .
the limit of the above equations that survives the extension to a large number of dirac fields also agrees with the previous calculation . @xcite the above flows , besides the gaussian , exhibit three fixed points at finite couplings ( fig .
the first critical point ( a ) is at @xmath210 , @xmath211 , and the second critical point ( c ) is at @xmath212 , @xmath213 .
there is also a bicritical fixed point ( b ) that separates the domains of attraction of the two critical points , at @xmath214 , @xmath215 .
the physical interpretation of the critical point a is obvious .
since we can tune through it by keeping @xmath211 and increasing @xmath201 over a certain negative value , it should describe the transition into the insulator with the gap that breaks the time - reversal symmetry , described by @xmath216 this state is obviously favored at a large and negative @xmath201 .
note that since @xmath20 commutes with @xmath50 , the line @xmath211 is invariant under rg .
in fact , the perturbative @xmath217-function along this line has to be identical as the one in the gross - neveu model .
we can therefore simply use the already existing higher - order estimates @xcite and the numerical results @xcite to find the critical exponents describing this particular metal - insulator transition .
we return to this fixed point shortly .
the physical interpretation of the critical point c is less obvious , but we can think of it as follows .
first , note that the line @xmath218 , @xmath219 , which describes the thirring model , @xcite belongs to the domain of attraction of c. also , the fierz transformations in eq .
( 50 ) imply @xmath220 the line @xmath221 for @xmath219 , which we name the nambu - jona - lasinio ( njl ) line , @xcite also falls into the domain of attraction of the critical point c. along this line , however , there should be a transition into an insulating state with @xmath222 where @xmath223 is a unit vector .
such a state is clearly favored at a large and positive @xmath202 along the njl line , and breaks the chiral @xmath3 symmetry down to @xmath224 .
we therefore identify c as the metal - insulator critical point governing the chiral - symmetry breaking transition in _ both _ thirring and njl models with a single dirac field .
the picture suggested by the above one - loop calculation in the maximally symmetric theory appears quite natural .
there are two possible insulating phases , each breaking either chiral or time - reversal symmetry , which correspond to possible `` masses '' for the dirac fermions .
both metal - insulator transitions are continuous , and are described by different critical points . of course , the true low - energy theory on the honeycomb lattice is much less symmetric than the one studied in this section .
nevertheless , we will argue that the two identified critical points may in fact be stable at least with respect to weak manifest breaking of the lorentz and chiral symmetries .
the explicit breaking of lorentz symmetry down to the rotational symmetry can be easily implemented by adding to @xmath225 a small symmetry breaking term @xmath226 which by virtue of being only rotational symmetric is guaranteed to be linearly independent of @xmath203 and @xmath204 .
a weak perturbation @xmath227 to the lowest order in couplings @xmath201 and @xmath202 then flows according to @xmath228 we thus find the critical point c to be stable with respect to weak lorentz symmetry breaking to one loop , the bicritical point b to be unstable , and a marginal .
we suspect that this result , although clearly an outcome here of an uncontrolled approximation , may be indicative of the true state of affairs .
hereafter we will assume that the critical points a and c are stable with respect to weak breaking of the lorentz symmetry in the lagrangian .
it may also be worth mentioning that the complete one - loop @xmath217-functions for @xmath227 , @xmath202 and @xmath201 , which we have computed but have not shown , do not lead to any new critical points at @xmath229 .
the simplest quartic term with the full lorentz symmetry , and only @xmath110 subgroup of the full chiral symmetry may be written as @xmath230 the fierz transformation matrix given above implies that these three quartic terms are indeed linearly independent .
when @xmath231 , the lagrangian @xmath196 acquires the full chiral @xmath232 symmetry , and may be rewritten as @xmath178 . using the fierz transformation , and after a convenient rescaling of the couplings as @xmath233 , to the quadratic order one finds @xmath234 @xmath235 @xmath236 the two chirally symmetric critical points from the previous section now appear at @xmath237 , @xmath238 ( a ) , and @xmath239 , @xmath240 ( c ) , and both remain critical , even in absence of chiral symmetry in the lagrangian . there is , however , an additional critical point ( e ) at @xmath241 , @xmath242 .
note also that the plane @xmath243 is invariant under the renormalization group , but whereas the fixed point a in that plane is critical , the fixed point ( d ) at @xmath218 , @xmath244 is bicritical .
one also finds that the line @xmath218 , @xmath245 , @xmath246-infinitesimal and positive intersects the critical surface which contains the point e , whereas for @xmath246-infinitesimal and negative the critical behavior is governed by c. for weak @xmath246 therefore there is a crossover from the fixed point at d toward either c or e , depending on the sign .
interestingly , for negative @xmath246 chiral symmetry becomes fully restored at the transition , at least within our one - loop calculation .
motivated by the one - loop results , we will assume hereafter that the lorentz symmetry becomes restored at long distances in the domain of interest , and that we need only consider @xmath190 with the three couplings from the last section .
the situation however , can then be simplified even further , as we discuss in this section .
consider the interaction hamiltonian in eq .
if the @xmath29-orbitals are well localized on their corresponding lattice sites , we may neglect the matrix elements with @xmath247 or @xmath248 . keeping only the remaining , dominant matrix elements
then one obtains the `` atomic limit '' of the general interaction hamiltonian @xmath249 where @xmath250 is the electron number operator at site @xmath72 .
the class of hamiltonians @xmath251 is evidently still rather broad , and would for example include all lattice interacting hamiltonians .
writing the lattice hamiltonian @xmath251 in terms of the dirac fields , however , imposes yet another restriction on the coupling constants . since
@xmath251 is written in terms of lattice - site particle number operators , any dirac quartic term evidently must contain an equal number of @xmath252 ( @xmath253 ) and @xmath32 ( @xmath33 ) fields . on the other hand ,
the @xmath254-term from above in momentum space can be written schematically as @xmath255 and thus contains the terms forbidden in the atomic limit,@xcite such as @xmath256 .
the index 1 and 2 refers here to the two dirac points .
this implies that for any lattice hamiltonian @xmath251 we must have @xmath257 note that the plane @xmath258 is invariant under the change of cutoff in the above one - loop calculation .
it is easy to see that this feature of the @xmath217-functions for @xmath201 , @xmath259 and @xmath246 is in fact true to all orders in perturbation theory .
the matrices @xmath20 and @xmath260 in the remaining two terms in @xmath196 commute with the dirac propagator , and therefore an arbitrary diagram containing @xmath201 and @xmath259 terms can contribute only to the renormalized @xmath201 and @xmath259 couplings .
so imposing @xmath258 at an arbitrary cutoff guarantees its vanishing at all others .
it is therefore not only physically justified but also internally consistent to consider only the two couplings @xmath201 and @xmath259 in the lorentz symmetric , but chirally asymmetric low - energy theory .
the one - loop result in this plane is depicted in fig .
the transition is either into the time - reversal - symmetry - broken , or into chiral - symmetry - broken insulator .
several features of this flow diagram that should be generally valid are worth mentioning .
\1 ) there should be two critical points , both unstable in a single direction .
bicriticality of a , for example , would imply that the transition for negative @xmath201 at a weak positive @xmath259 is first order .
this , however , seems unlikely on physical grounds , and , also , it is not found in the explicit large - n generalization of the theory , when the two @xmath217-functions are known to decouple .
\2 ) the two critical points have identical critical behavior .
this is because the term @xmath261 , in graphene representation , under the transformation @xmath262 \psi,\ ] ] @xmath263,\ ] ] goes into @xmath264 , and vice versa , while @xmath47 remains invariant .
this also means that the two @xmath217-functions are symmetric under the exchange @xmath265 .
both critical points are thus in the universality class of the gross - neveu model .
\3 ) at the line @xmath266 the single @xmath217-function becomes @xmath267 i. e. @xmath201 flows according solely to its canonical dimension .
this is because @xmath268 where @xmath269 . since all @xmath50 are block - diagonal , @xmath270 and @xmath271 components at this line decouple .
the partition function factorizes into a product of two gross - neveu partition functions , each containing a single _ two - component _ dirac fermion . along this line
the system is believed to have the metal - insulator transition , possibly continuous , @xcite but the @xmath217-function vanishes at least to the order @xmath272 .
each identified metal - insulator transition is characterized by a set of critical exponents .
we will here focus on the three already mentioned in the introduction : the correlation length exponent @xmath273 , the dynamical exponent @xmath25 , and the dirac fermion anomalous dimension @xmath28 .
the other exponents can then be obtained from the usual scaling relations .
@xcite first , since all the identified critical points exhibit lorentz symmetry , @xmath274 we also find that the exponent @xmath273 is unity at all critical points as well , but this is clearly an artifact of the one - loop calculation . in general
, @xmath273 is expected to be different at different critical points .
the same goes for @xmath28 , which vanishes in one - loop calculation , but will be finite in general . in the atomic limit , when under the assumed lorentz invariance we need only two coupling constants , the values of the critical exponents are better known .
first , in a perturbative calculation in powers of coupling constants , the exponents at critical points @xmath275 and @xmath276 will be identical .
we may thus expect that in a lattice theory with short - range repulsion the transition is either into the time - reversal symmetry or chiral symmetry broken insulator , in either case with @xcite @xmath277 @xmath278 note that since there is only a single dirac field involved the numerical values of the exponents @xmath273 and @xmath28 differ significantly from the large - n values of unity and zero , respectively .
this raises hope that this non - trivial critical behavior may be observable in numerical simulations of lattice models . although not of immediate relevance to graphene
, it would still be of interest to determine the critical exponents at the chirally symmetric critical point c , which we proposed to control the critical behavior of the thirring and the njl model . we , however , are not aware of any analytical nor numerical study of the njl model that goes beyond the leading order in large - n calculation .
the critical exponents , as usual , govern for example the critical behavior of the gap on the insulating side of the transition , as mentioned in the introduction . in the present case , however , there are massless fermionic excitations on the metallic side , and one may wonder if and how the approach to the critical point is reflected onto these .
let us therefore generalize slightly and provide the support for the results already announced in the introduction .
first , the usual scaling @xcite implies that at the cutoff @xmath279 electron s two - point correlation function near the critical point and at zero temperature satisfies @xmath280 where @xmath281 is the matsubara frequency and @xmath282 is the transition s tuning parameter . setting @xmath283 we thus find the usual scaling law @xmath284 where @xmath285 is a universal scaling function . from here
we can extract the scaling of the fermi velocity and quasiparticle residue as follows .
first , if upon the analytical continuation to real frequencies @xmath286 has a pole at @xmath287 the scaling relation immediately dictates that @xmath288 let us next set @xmath289 , take @xmath290 , and let @xmath291 . in this limit @xmath292 and therefore @xmath293 . in order to cancel the @xmath294-dependence of the prefactor in eq .
( 85 ) in this limit it must be that @xmath295 where @xmath296 is the ( momentum ) anomalous dimension .
analogously , assuming that for @xmath297 , @xmath298 , one finds that also @xmath299 on the other hand , in the opposite limit @xmath300 and @xmath301 , in the metallic phase at low energies we have fermionic quasiparticles .
this implies that , for example , @xmath302 , i. e. @xmath303 with @xmath304 . combining with the previous relation
, the quasiparticle pole s residue behaves as @xmath305 for @xmath17 , the two anomalous dimensions are the same , @xmath306 , where @xmath307 is the dirac fermion s anomalous dimension , and the scaling announced in the introduction follows .
the special form of this relation for large number of dirac components when besides @xmath17 , it is also @xmath22 , was previously proposed .
@xcite the quasiparticle residue vanishes upon the approach to the metal - insulator transition , as proposed long - ago by brinkman and rice , but here in a decidedly non - mean - field fashion .
there are at least two obvious generalizations important for real graphene : the addition of spin , and the inclusion of the long - range tail of coulomb repulsion . adding spin
would simply double the number of couplings for each symmetry , since each independent quartic term would then require a separate coupling in the singlet and in the triplet channels .
the minimal internally consistent low - energy theory would then be the generalization of the lorentz invariant lagrangian in eq .
( 70 ) , with @xmath258 : @xmath308 where the sum goes over @xmath309 , and @xmath310 , with @xmath311 , and @xmath312 .
the lagrangian with @xmath313 would represent the extended hubbard model with on - site and nearest neighbor repulsion , considered before in the limit of large number of dirac fermions in ref .
the interplay between the various instabilities in the theory equivalent to the above lagrangian was recently studied in @xcite , where it was pointed out that the second - nearest - neighbor repulsion implies a negative coupling @xmath314 for example .
the form of the above minimal spinful lagrangian facilitates a systematic study of the metal - insulator transition in the hubbard model , which will be a subject of a separate publication .
few comments on the importance of long - range tail of coulomb interaction are also in order
. weak coulomb ( @xmath315 ) interaction is an ( marginally ) irrelevant perturbation at the gaussian fixed point , and this remains true at the metal - insulator critical point at large - n as well @xcite .
furthermore , the entire @xmath217-function for the charge coupling @xmath316 can be computed at large - n , and it does not exhibit any non - trivial zeroes .
@xcite on the other hand , several calculations show that by increasing the coupling @xmath316 beyond certain point and for small enough @xmath317 the system can be tuned through a metal - insulator transition at which the chiral symmetry becomes spontaneously broken.@xcite the nature of such a putative metal - insulator transition is in our mind an open question at the moment .
whereas it is possible that it is described by new `` charged '' critical point @xcite corresponding to the non - trivial zero of the @xmath318 , it also seems conceivable that the charge is always irrelevant and that the transition is still in the universality class of the critical point c , in our nomenclature .
yet another possibility is a discontinuous transition .
more work is obviously needed in order to be able to address this issue more conclusively .
it may also be interesting to note that in the related bosonic problem , when a systematic expansion near four dimension is readily available , there are no charged critical points in the theory to the leading order .
@xcite this may also be contrasted with the well - known example of ( albeit lorentz invariant ) scalar higgs electrodynamics , for which the critical points , when they exist , are always charged .
@xcite probably the central message of this work is that provided lorentz invariance becomes emergent near criticality , for the @xmath29-orbitals well localized on carbon atoms the lagrangian may be taken to contain only two ( or with the physical spin , four ) coupling constants .
if there are no intervening first - order transitions , one can infer then that there are two possible continuous metal - insulator transitions , both governed by the same gross - neveu model in 2 + 1 dimensions , into the states that break either time - reversal or chiral symmetry .
the residue of the quasiparticle pole on the metallic side plays the role of the metal s order parameter , and it vanishes continuously with a small critical exponent proportional to the fermion s anomalous dimension .
in contrast , the specific heat coefficient @xmath319 being dependent on the fermi velocity only , at the transition vanishes discontinuously from a finite value on the metallic side .
near the critical point and for the temperatures much below the bandwidth we may assume that the specific heat obeys the scaling relation @xmath320 for small arguments the universal scaling function @xmath321 behaves as @xmath322 , so that in the metallic phase one finds the usual quadratic temperature dependence @xmath323 recognizing the proportionality of the specific heat coefficient as @xmath324 gives us yet another way to deduce eq .
( 87 ) . at the criticality , on the other hand , @xmath325 with @xmath326 expected to be finite .
when @xmath17 , the specific heat coefficient near criticality jumps therefore from @xmath327 in the metallic phase to @xmath326 at the critical point , and finally to zero in the insulating phase.@xcite similarly , the optical conductivity near the metal - insulator transition will obey the scaling relation for @xmath300 @xmath328 with @xmath329 as a universal function , and with @xmath330 as the familiar universal dc conductivity per dirac field in the metallic phase .
@xcite in contrast to the specific heat , there is no ( non - universal ) dimensionful quantity such as @xmath331 in the scaling expression for conductivity , and consequently @xmath332 is constant and universal in the entire metallic phase .
right at the transition then @xmath333
so the dc conductivity , while still universal , at the criticality should be different than in the metallic phase .
finally , in the insulator the dc conductivity vanishes , so that the dc conductivity , similar to the specific heat coefficient , in principle should show two universal discontinuities at the metal - insulator transition .
one obstacle to experimental observation of these predictions is that it has not been possible yet to tune the parameter @xmath334(interaction / bandwidth ) in graphene , and sample different phases of the system .
the application of the magnetic field , however , changes this , since the kinetic energy becomes completely quenched , and infinitesimal interaction immediately induces a finite gap . if the parameters of the system place it not too far from the metal - insulator transition , the gap @xmath335 would obey @xcite @xmath336 where @xmath337 is the magnetic length , @xmath338 is the lattice constant , and @xmath294 is the tuning parameter .
@xmath339 is a ( universal ) scaling function .
the computation of the scaling function in the large - n limit , and the consequences of this scaling relation for experiment were discussed at length before .
@xcite here we only wish to underline that the emergent lorentz invariance of the metal - insulator critical point , via its consequence that @xmath17 , implies precise proportionality between the interaction gap and the landau level separation at the criticality , @xmath340 where @xmath341 is a universal number .
such a square - root magnetic field dependence of the gap is well known to arise from the long - range tail of the coulomb interactions , but the above derivation serves to show that its origin may in principle lie in purely short - range interactions as well .
we have presented the theory of electrons interacting via short - range interactions on honeycomb lattice , and in particular , determined the number and types of independent quartic terms in the low - energy lagrangian .
metal - insulator quantum critical points and the concomitant quantum critical behavior were discussed , with the particular attention paid to the consequences of the emergent lorentz invariance .
the minimal internally consistent local lagrangian for spinless fermions is shown to contain only two gross - neveu - like quartic terms .
generalizations that would include long - range coulomb interaction or spin of electrons were briefly considered .
we also discussed the critical behavior of several key physical quantities on the metallic side of the transition , such as the fermi velocity , the residue of the quasiparticle pole , specific heat , and the frequency dependent conductivity .
this work was supported by the nserc of canada .
for completeness , we provide the derivation of the fierz identity . assume a basis @xmath342 in the space of four - dimensional matrices , and choose @xmath343 .
then any hermitean matrix @xmath344 can be written as @xmath345 with the summation over repeated indices assumed .
this can be rewritten as @xmath346 and therefore it follows that @xmath347 applying this identity to the product of two matrix elements then yields @xmath348 finally , this leads to the expansion of a quartic term as @xmath349 which is used in the text for @xmath350 .
the minus sign in the last line derives from the grassmann nature of the fermionic fields .
any real @xmath317-dimensional matrix @xmath344 can obviously be written as @xmath351 where @xmath352 , and @xmath353 . in dirac notation , @xmath354 and @xmath355 is the transposed matrix
. there exists such a representation of the matrix @xmath344 in any basis of vectors @xmath356 , as can be seen by multiplying @xmath344 from the right with @xmath357 .
let us now form a related symmetric matrix @xmath358 .
being symmetric , it can be written in the usual spectral form @xmath359 where @xmath360 .
we can now write , however , the matrix @xmath344 in the particular eigenbasis of the associated symmetric matrix @xcite @xmath361 @xmath362 from the definition of @xmath84 and its spectral form we see that @xmath363 , and therefore @xmath364 where @xmath365 , and @xmath366 . for a general asymmetric matrix , the basis @xmath367 and @xmath273 are different , and the last equation generalizes the more familiar form for a symmetric matrix , where they are the same .
here we provide an alternative formulation of the renormalization group transformation in presence of constraints imposed by the fierz identity .
let us demonstrate this method on the simplest example of the maximally symmetric theory . instead of choosing two independent couplings and using fierz transformation at intermediate stages of the calculation to transform any other generated quartic terms back into the chosen ones , one may use the kernel of the fierz matrix to write the lagrangian in terms only of the physical couplings from the outset , as in eq .
the advantage of doing this is that no other quartic term besides the ones corresponding to the physical couplings can ever get generated then by the renormalization transformation .
the set of couplings @xmath188 and @xmath189 is therefore closed under renormalization .
the computation to the quadratic order then yields @xmath368 @xmath369 the connection to eqs .
( 63)-(64 ) in the text can be established as follows . since the fierz transformations in this case imply @xmath370 @xmath371 the lagrangian @xmath178 in eq .
( 41 ) can obviously also be written as @xmath372 in the text we therefore have simply named the entire first bracket @xmath201 , and the second @xmath202 . but these can be recognized as particular linear combinations of the physical couplings @xmath188 and @xmath189 : @xmath373 @xmath374 such a connection is of course completely general , and in particular may be established between the three chosen couplings in the eq .
( 70 ) and the `` physical couplings '' determined by the vectors in eqs .
( 58)-(60 ) .
99 for reviews , see v.p .
gusynin , s.g .
sharapov , j.p .
carbotte , int .
j. of mod .
b , * 21 * , 4611 ( 2007 ) ; a. h. castro neto , f. guinea , n. m. r. peres , k. s. novoselov , a. k. geim , rev .
phys . * 81 * , 109 ( 2009 ) .
i. f. herbut , phys .
. lett . * 97 * , 146401 ( 2006 ) .
j. gonzalez , f. guinea , m. a. h. vozmediano , nucl .
b424 * , 595 ( 1994 ) ; phys .
b * 59 * , 2474 ( 1999 ) .
o. vafek , phys .
98 * , 216401 ( 2007 ) .
i. f. herbut , phys .
lett . * 99 * 206404 , ( 2007 ) .
s. sorella and e. tosatti , europhys . lett .
* 19 * , 699 ( 1992 ) .
l. m. martelo , m. dzierzawa , l. siffert , and d. baeriswyl , z. phys .
b * 103 * , 335 ( 1997 ) . t. paiva , r. t. scalettar , w. zheng , r. r. singh , and j. oitmaa , phys . rev .
b * 72 * , 085123 ( 2005 ) .
y. zhang , z. jiang , j. p. small , m. s. purewal , y .- w .
tan , m. fazlollahi , j. d. chudow , j. a. jaszczak , h. l. stormer , and p. kim , phys .
lett . * 96 * , 136806 ( 2006 ) ; z. jiang , y. zhang , h. l. stormer , and p. kim , phys .
lett . * 99 * , 106802 ( 2007 ) .
joseph g. checkelsky , lu li , n. p. ong , phys .
* 100 * , 206801 ( 2008 ) ; preprint arxiv:0808.0906 .
d. v. khveshchenko , phys .
rev . lett . * 87 * , 206401 ( 2001 ) ; ibid .
* 87 * , 246802 ( 2001 ) .
v. p. gusynin , v. a. miransky , s. g. sharapov , and i. a. shovkovy , phys .
b * 74 * , 195429 ( 2006 ) ; e. v. gorbar , v. p. gusynin , v. a. miransky , preprint arxiv:0710.3527 . i. f. herbut , phys . rev .
b * 75 * , 165411 ( 2007 ) ; ibid .
* 76 * , 085432 ( 2007 ) ; i. f. herbut and b. roy , * 77 * , 245438 ( 2008 ) .
j. alicea and m. p. a. fisher , phys .
b * 74 * , 075422 ( 2006 ) ; r. l. doretto and c. morais - smith , phys .
b * 76 * , 195431 ( 2007 ) . for a review ,
see a. k. geim and k. s. novoselov , nature materials * 6 * , 183 ( 2007 ) .
i. f. herbut , v. jurii ' c , and o. vafek , phys .
lett . * 100 * 046403 ( 2008 ) ; i. f. herbut , v. jurii ' c , o. vafek , m. j. case , preprint , arxive:0809.0725 .
l. fritz , j. schmalian , m. mueller , and s. sachdev , phys .
b * 78 * , 085416 ( 2008 ) . by this definition
the interaction that at large particle separation decays as a power law @xmath375 is short ranged provided that @xmath376 .
see for example , y. takahashi , in _ progress in quantum field theory _
, ed . by h. ezawa and s. kamefuchi ( north holland , 1986 ) . for the closely related chiral symmetry of d - wave superconductors , see i. f. herbut , phys .
b * 66 * , 094504 ( 2002 ) ; phys .
lett . * 88 * , 047006 ( 2002 ) ; phys .
* 94 * , 237001 ( 2005 ) ; d. j. lee and i. f. herbut , phys .
b * 66 * , 094512 ( 2002 ) ; b. seradjeh and i. f. herbut , phys . rev .
b * 66 * , 184507 ( 2002 ) ; z. teanovi ' c , o. vafek , and m. franz , phys . rev .
b * 65 * , 180511 ( 2002 ) ; m. franz , t. pereg - barnea , d.e . sheehy , and z. teanovi ' c , phys .
b * 68 * , 024508 ( 2003 ) ; i. o. thomas and s. hands , phys . rev .
b * 75 * , 134516 ( 2007 ) . at the critical point at a finite coulomb interaction in two - dimensions @xmath25 is always unity ; see i. f. herbut , phys .
lett . * 87 * , 137004 ( 2001 ) .
f. d. m. haldane , phys .
lett . * 61 * , 2015 ( 1988 ) .
g. w. semenoff , phys .
lett . * 53 * , 2449 ( 1984 ) .
hou , c. chamon , and c. mudry , phys .
* 98 * , 186809 ( 2007 ) .
i. herbut , _ a modern approach to critical phenomena _ , ( cambridge university press , cambridge , 2007 ) .
w. f. brinkman and t. m. rice , phys .
b * 2 * , 4302 ( 1970 ) .
y. hasegawa , r. konno , h. nakano , m. kohmoto , phys .
b * 74 * , 033413 ( 2006 ) .
i. f. herbut , phys .
b. * 78 * , 205433 ( 2008 ) .
muicki and b. mili ' c , _ matematicke osnove teorijske fizike _ , ( univerzitet u beogradu , beograd , 1984 ) , sec . 5.3 ( in serbian ) . here
we treat the frequency integrals in the loops analogously to momentum integrals .
for an alternative , see ref .
2 , for example .
k. kaveh and i. f. herbut , phys .
b * 71 * , 184519 ( 2005 ) .
a. m. vasilev , s. e. derkachov , n. a. kilev , and a. s. stepanenko , teor . mat .
fiz . * 92 * , 486 ( 1992 ) ; _ ibid . _ * 97 * , 364 ( 1993 ) ; j. a. gracey , int . j. mod . phys . a * 9 * , 727 ( 1994 ) .
l. krkkinen , l. lacaze , p. lacock , and b. petersson , nucl .
b * 415 * , 781 ( 1994 ) ; * 438 * 650(e ) ( 1995 ) .
l. rosa , p. vitale , and c. wetterich , phys .
* 86 * , 958 ( 2001 ) ; f. hfling , c. novak , and c. wetterich , phys .
b * 66 * , 205111 ( 2002 ) .
s. christofi , s. hands , c. strouthos , phys.rev .
d * 75 * , 101701 ( 2007 ) , and references therein .
y. nambu and g. jona - lasinio , phys . rev . *
122 * , 345 ( 1961 ) . in the purely bosonic @xmath377-theories ,
the @xmath378 symmetry is currently believed not to emerge at the critical point out of @xmath379 , in three dimensions .
the critical behavior is governed by the `` biconal '' fixed point , which however , appears to be extremely close the @xmath378-symmetric point in the coupling space .
see , p. calabrese , a. pelissetto , e. vicari , phys .
b * 67 * , 054505 ( 2003 ) .
the terms allowed in the atomic limit are already contained in the terms @xmath203 and @xmath380 .
see , nevertheless , @xcite for numerical results on thirring models with more than one dirac fermion . s. raghu , xiao - liang qi , c. honerkamp , s .- c .
zhang , phys .
lett . * 100 * , 156401 ( 2008 ) .
i. l. aleiner , d. e. kharzaev , and a. m. tsvelik , phys .
b * 76 * , 195415 ( 2007 ) d. t. son , phys . rev .
b * 75 * , 235423 ( 2007 ) ; j. e. drut and d. t. son , phys . rev .
b * 77 * , 075115 ( 2008 )
. e. v. gorbar , v. p. gusynin , v. a. miransky , and i. a. shovkovy , phys . rev .
b * 66 * , 045108 , ( 2002 ) .
d. v. khveshchenko and h. leal , nucl . phys .
b * 687 * , 323 ( 2004 ) .
j. e. drut and t. a. lahde , phys .
lett , * 102 * , 026802 ( 2009 ) . s. hands and c. strouthos , phys
b * 78 * 165423 , 2008 . o. vafek and m. j. case , phys .
b * 77 * , 033410 ( 2008 ) . see the ref . 21 , and the problem 8.10 in ref .
i. f. herbut and z. teanovi ' c , phys .
* 76 * , 4588 ( 1996 ) ; ibid . *
78 * , 980 ( 1997 ) ; i. f. herbut , j. phys .
a : math . gen . * 30 * , 423 ( 1997 ) . for spinless fermions considered here
, both insulators break the discrete ising symmetry , so there are no nambu - goldstone bosons . | the general low - energy theory of electrons interacting via repulsive short - range interactions on graphene s honeycomb lattice at half filling is presented .
the exact symmetry of the lagrangian with local quartic terms for the dirac four - component field dictated by the lattice is identified as @xmath0time reversal , where @xmath1 is the dihedral group , and @xmath2 is a subgroup of the @xmath3 `` chiral '' group of the non - interacting lagrangian , that represents translations in dirac language .
the lagrangian describing spinless particles respecting this symmetry is parameterized by six independent coupling constants .
we show how first imposing the rotational , then lorentz , and finally chiral symmetry to the quartic terms - in conjunction with the fierz transformations - eventually reduces the set of couplings to just two , in the `` maximally symmetric '' local interacting theory .
we identify the two critical points in such a lorentz and chirally symmetric theory as describing metal - insulator transitions into the states with either time - reversal or chiral symmetry being broken .
the latter is proposed to govern the continuous transition in both the thirring and nambu - jona - lasinio models in 2 + 1 dimensions and with a single dirac field . in the site - localized , `` atomic '' , limit of the interacting hamiltonian , under the assumption of emergent lorentz invariance
, the low - energy theory describes the continuous transitions into the insulator with either a finite haldane s ( circulating currents ) or semenoff s ( staggered density ) masses , both in the universality class of the gross - neveu model .
the simple picture of the metal - insulator transition on a honeycomb lattice emerges at which the residue of the quasiparticle pole at the metallic , and the mass - gap in the insulating phase both vanish continuously as the critical point is approached .
in contrast to these two critical quantities , we argue that the fermi velocity is non - critical as a consequence of the dynamical exponent being fixed to unity by the emergent lorentz invariance near criticality .
possible effects of the long - range coulomb interaction , and the critical behavior of the specific heat and conductivity are discussed . |
due to its multifaceted roles in nuclear physics and astrophysics @xcite as well as new physics beyond the standard model @xcite , the symmetry energy has become a hot topic in current research frontiers of nuclear physics and astrophysics @xcite . during the last decade , a lot of experimental , observational and theoretical efforts have been devoted to constraining the magnitude @xmath7 and density slope @xmath8 of the symmetry energy at nuclear saturation density @xmath9 ( @xmath10 @xmath5 ) , i.e. , @xmath11 and @xmath12 . although important progress has been made , large uncertainties on the values of @xmath11 and @xmath12 still exist ( see , e.g. , refs .
for instance , while the @xmath11 is determined to be around @xmath13 mev , the extracted @xmath12 varies significantly from about @xmath14 to @xmath15 mev , depending on the observables and analysis methods . to better understand the model dependence and narrow the uncertainties of the constraints is thus of extreme importance . while many studies on heavy ion collisions and neutron stars have significantly improved our knowledge on the symmetry energy , more and more constraints on the symmetry energy have been obtained in recent years from analyzing the properties of finite nuclei , such as the nuclear binding energy @xcite , the neutron skin thickness @xcite , and the resonances and excitations @xcite .
furthermore , it has been realized that the properties of finite nuclei usually provide more precise constraints on @xmath7 and @xmath8 at subsaturation densities rather than at saturation density @xmath9 .
this feature is understandable since the characteristic ( average ) density of finite nuclei is less than @xmath9 . for example , the average density of heavy nuclei ( e.g. , @xmath1pb ) is about @xmath16 @xmath5 , and thus the properties of heavy nuclei most effectively probe the properties of nuclear matter around @xmath16 @xmath5 @xcite .
indeed , a quite accurate constraint on the symmetry energy at the subsaturation cross density @xmath4 @xmath5 , i.e. , @xmath17mev , has been recently obtained from analyzing the binding energy difference of heavy isotope pairs @xcite .
in contrast to the fact that many and precise constraints on the magnitude of @xmath18 around @xmath19 have been obtained , to the best of our knowledge , so far there is only one experimental constraint on the density slope @xmath3 which was obtained from analyzing the neutron skin data of sn isotopes @xcite .
knowledge on @xmath3 is not only important for understanding the density dependence of the symmetry energy itself , but also plays a central role in determining the neutron skin thickness of heavy nuclei and the core - crust transition density in neutron stars .
therefore , any new constraints on @xmath3 will be extremely useful . in the present work , with the precise knowledge of @xmath2
, we demonstrate that the electric dipole polarizability @xmath20 in @xmath1pb measured at the research center for nuclear physics ( rcnp ) via polarized proton inelastic scattering at forward angles , can put a strong limit on the @xmath3 .
we emphasize since at forward angles coulomb excitation dominates , the extracted @xmath20 at rcnp is expected to be a relatively clean isovector indicator with less uncertainties from strong interaction .
the equation of state ( eos ) of asymmetric nuclear matter , given by its binding energy per nucleon , can be written as @xmath21where @xmath22 is the baryon density , @xmath23 is isospin asymmetry , @xmath24 is the eos of symmetric nuclear matter , and the symmetry energy is expressed as @xmath25around a reference density @xmath26 , the @xmath27 can be expanded in @xmath28 as @xmath29 where @xmath30 is the density slope parameter which characterizes the density dependence of the symmetry energy around @xmath26 .
our calculations in the present work are based on the skyrme - hartree - fock ( shf ) approach with the so - called standard skyrme force ( see , e.g. , ref .
@xcite ) which includes @xmath31 parameters , i.e. , the @xmath32 skyrme force parameters @xmath33 , @xmath34 , @xmath35 , and the spin - orbit coupling constant @xmath36 . instead of directly using the @xmath32 skyrme force parameters
, we can express them explicitly in terms of @xmath32 macroscopic quantities , i.e. , @xmath37 , @xmath38 , the incompressibility @xmath39 , the isoscalar effective mass @xmath40 , the isovector effective mass @xmath41 , @xmath42 , @xmath43 , the gradient coefficient @xmath44 , and the symmetry - gradient coefficient @xmath45 . in this case
, we can examine the correlation of properties of finite nuclei with each individual macroscopic quantity by varying individually these macroscopic quantities within their empirical ranges .
recently , this correlation analysis method has been successfully applied to study nuclear matter properties from analyzing nuclear structure observables @xcite , and will also be used in this work .
the random - phase approximation ( rpa ) provides an important microscopic approach to calculate the electric dipole polarizability in finite nuclei . within the framework of rpa theory , for a given excitation operator @xmath46 , the reduced transition probability from rpa ground state @xmath47 to rpa excitation state @xmath48
is given by : @xmath49 where @xmath50 denotes the unoccupied ( occupied ) single nucleon state ; @xmath51 is the reduced matrix element of @xmath46 ; and @xmath52 and @xmath53 are the rpa amplitudes .
the strength function then can be calculated as : @xmath54 where @xmath55 is the energy of rpa excitation state @xmath56 .
thus the moments of strength function can be obtained as : @xmath57 in the case of electric dipole ( @xmath58 ) response , the excitation operator is defined as : @xmath59 where @xmath60 , @xmath61 and @xmath62 are proton , neutron and mass number , respectively ; @xmath63 is the nucleon s radial coordinate ; @xmath64 is the corresponding spherical harmonic function . for a given skyrme interaction
, we can calculate the inverse energy - weighted moment @xmath65 using the hf - rpa method , and then obtain the electric dipole polarizability @xmath66 as @xmath67 for the theoretical calculations of electric dipole polarizability in @xmath1pb in the present work , we employ the skyrme - rpa program by col@xmath68 _
@xcite . in this program , the shf mean field and
the rpa excitations are fully self - consistent .
in particular , we calculate the isovector dipole strength in @xmath1pb with a spherical box extending up to @xmath69 fm , a radial mesh of @xmath70 fm and a cutoff energy of @xmath71 mev which denotes the maximum energy of the unoccupied single - particle states in the rpa model space . then the inverse energy - weighted moment @xmath65 is evaluated with an upper integration limit of @xmath72 mev according to the experimental energy range @xcite , and thus the electric dipole polarizability @xmath20 can be calculated invoking eq . .
the electric dipole polarizability @xmath20 has been shown to be a sensitive probe of the symmetry energy @xcite .
in particular , based on the droplet model , roca - maza _ et al . _
@xcite obtained the following relation : @xmath73,\ ] ] where @xmath74 is the mean - square radius and @xmath75 is the symmetry energy coefficient of a finite nucleus of mass number @xmath62 .
furthermore , using the empirical relation @xmath76 @xcite and expanding @xmath77 as @xmath78 roca maza _
et al . _ demonstrated that @xmath79 is correlated with both @xmath80 and @xmath12 .
particularly , based on a large and representative set of relativistic and nonrelativistic nuclear mean - field models , they found a strong linear correlation between @xmath81 and @xmath12 and then extracted the constraint @xmath82 mev from the combination of the experimental determination of @xmath79 with the empirical estimate of @xmath83 mev .
one can see that the uncertainty of the estimated @xmath80 leads to a large error of @xmath84 mev for @xmath12 . instead of expressing @xmath77 in terms of @xmath80 and @xmath12 as in eq .
( [ esymra ] ) , one can also express @xmath80 in terms of @xmath85 and @xmath86 as @xmath87 noting @xmath88 @xcite , one can then see from eqs .
( [ alphaddm ] ) and ( [ esymr0rc ] ) that @xmath79 in @xmath1pb is also correlated with both @xmath86 and @xmath85 . as we will see in the following , the microscopic rpa calculations indeed show that @xmath89 is sensitive to @xmath80 and @xmath12 as well as to @xmath86 and @xmath85 .
since @xmath85 has been stringently constrained recently ( see , e.g. , @xmath90 mev in ref.@xcite ) , the @xmath79 in @xmath1pb can thus be used to constrain the @xmath86 parameter .
to examine the correlation of the @xmath20 in @xmath1pb with each macroscopic quantity , especially on @xmath91 and @xmath92 , we show in fig .
[ alphad ] the @xmath20 in @xmath1pb from shf with the skyrme force msl0 @xcite by varying individually @xmath92 , @xmath45 , @xmath44 , @xmath93 , @xmath94 , @xmath39 , @xmath40 , @xmath41 , @xmath37 , and @xmath36 within their empirical uncertain ranges , namely , varying one quantity at a time while keeping all others at their default values in msl0 , for @xmath95 and @xmath96 @xmath5 , respectively .
it is seen from fig .
[ alphad ] that , as eq .
( [ alphaddm ] ) suggests , the @xmath66 in @xmath1pb exhibits strong correlations with both @xmath97 and @xmath98 , while much weaker correlation with other macroscopic quantities .
particularly , the @xmath20 decreases sensitively with @xmath99 while increases rapidly with @xmath100 , implying a fixed value of @xmath20 will lead to a strong positive correlation between @xmath99 and @xmath100 .
the results for @xmath101 @xmath5 just confirm the correlations of @xmath79 with @xmath80 and @xmath12 reported in ref.@xcite .
for @xmath95 @xmath5 , given that the symmetry energy at @xmath102 @xmath5 has been well constrained as @xmath103 mev , one thus expects the @xmath20 in @xmath1pb can constrain stringently the parameter @xmath104 . fixing the values of other @xmath105 macroscopic quantities , i.e. , @xmath45 , @xmath44 , @xmath93 , @xmath39 , @xmath40 , @xmath41 , @xmath37 and @xmath36 at their default values in msl0
, we illustrate in fig .
[ alphadlc ] by open up - triangles ( down - triangles ) the @xmath20 in @xmath1pb as a function of @xmath3 for @xmath106 mev .
as expected , it is seen from fig .
[ alphadlc ] that the @xmath20 in @xmath1pb increases ( decreases ) with @xmath104 ( @xmath2 ) for a fixed @xmath107 ( @xmath104 ) . by comparing with the experimental data @xmath108 @xmath109
, one can extract a strong constraint of @xmath110 mev . in @xmath1pb as a function of @xmath3 for fixed @xmath2 .
the open ( solid ) up- and down - triangles represent the results with @xmath111 and @xmath112 mev , respectively , from shf - rpa calculations with the values of other parameters fixed in msl0 ( obtained in optimization ) .
the band indicates the experimental value of @xmath113 @xmath109 from rcnp @xcite . ]
the above constraint of @xmath110 mev has been obtained by neglecting the weak correlations between the @xmath20 in @xmath1pb and other @xmath105 macroscopic quantities . to test the robustness of this constraint and to obtain a more precise constraint , for fixed
@xmath2 and @xmath3 , we optimize all other @xmath105 parameters instead of simply fixing them at their default values in msl0 , by minimizing the weighted sum of @xmath114 evaluated from the difference between shf prediction and the experimental data for some selected observables using the simulated annealing technique @xcite . in particular , in the optimization , we chose the following experimental data of spherical even - even nuclei , i.e. , ( i ) the binding energy @xmath115 of @xmath116o,@xmath117ca , @xmath118ni , @xmath119sr , @xmath120zr , @xmath121sn , @xmath122sm , @xmath1pb @xcite ; ( ii ) the charge rms radii @xmath123 of @xmath116o , @xmath117ca , @xmath124ni , @xmath119sr , @xmath120zr , @xmath125sn , @xmath122sm , @xmath1pb @xcite ; ( iii ) the breathing mode energy @xmath126 of @xmath120zr,@xmath127sn,@xmath122sm and @xmath1pb @xcite . in the calculation of the breathing mode energy @xmath128
, we evaluate the inverse energy - weighted sum rule @xmath65 with the constrained hartree - fock ( chf ) method and obtain the energy - weighted sum rule @xmath129 using the double commutator sum rule @xcite .
in addition , in the optimization , we constrain the macroscopic parameters by requiring that ( i ) the neutron @xmath130 energy level splitting in @xmath1pb should lie in the range of @xmath131 mev ; ( ii)@xmath132 should be greater than @xmath133 and here we set @xmath134 ( @xmath135 is nucleon mass in vacuum ) to be consistent with the extraction from global nucleon optical potentials constrained by world data on nucleon - nucleus and ( p , n ) charge - exchange reactions @xcite . as usual , in the optimization , we assign a theoretical error @xmath136 mev to e@xmath137 , @xmath138 fm to @xmath139 while use the experimental error for breathing mode energy @xmath140 with a weight factor @xmath141 , so that the respective @xmath142 evaluated from each sort of experimental data is roughly equal to the number of the corresponding data points @xcite . using the above optimization process , we evaluate the electric dipole polarizability @xmath20 in @xmath1pb as a function of @xmath3 for a fixed @xmath2 , and the results are shown in fig [ alphadlc ] by solid up - triangles ( down - triangles ) for @xmath106 mev . it should be noted that for each pair of @xmath2 and @xmath3 with fixed values , the other @xmath105 macroscopic quantities have been optimized accordingly as described above .
it is interesting to see that the values of @xmath20 with optimization are quite consistent with the results using the default values in msl0 without optimization and only show a small upward shift compared with the latter . comparing the results from optimization to the experimental data , one can obtain a constraint of @xmath6 mev , which is again in good agreement with the constraint @xmath110 mev extracted using the default values in msl0 .
these features demonstrate the validity of neglecting the weak correlations between the @xmath20 in @xmath1pb and other @xmath105 macroscopic quantities .
the present constraint on @xmath3 further agrees very well with the constraint @xmath143 mev extracted from analyzing the experimental data on the neutron skin thickness of sn isotopes @xcite .
this is a very interesting finding since these two constraints are obtained from two completely independent experimental observables .
in addition , using the constrained @xmath2 and @xmath3 together with the corresponding @xmath105 other optimized quantities , one can easily extract the @xmath7 and @xmath144 at saturation density @xmath9 , and the results are @xmath145 mev and @xmath146 mev , which are essentially consistent with other constraints extracted from terrestrial experiments , astrophysical observations , and theoretical calculations with controlled uncertainties @xcite . especially , our present results agree surprisingly well with the constraint of @xmath147-@xmath148 mev and @xmath149-@xmath150 mev ( at 95% confidence level ) obtained from analyzing the mass and radius of neutron stars @xcite as well as that of @xmath151-@xmath152 mev and @xmath153-@xmath154 mev extracted from the experimental , theoretical and observational analyses @xcite .
our results are also in agreement with the constraint of @xmath155 mev and @xmath156 mev from analyzing pygmy dipole resonances ( pdr ) of @xmath157sn @xcite and that of @xmath158 mev and @xmath159 mev from analyzing pdr of @xmath160ni and @xmath161sn @xcite . in addition , our results are further consistent with the constraint of @xmath162 mev and @xmath163 mev extracted from analyzing the experimental data of the binding energy difference of heavy isotope pairs and the neutron skins of sn isotopes @xcite as well as the constraint of @xmath164 mev and @xmath165 mev from a new finite - range droplet model analysis of the nuclear mass @xcite . given that the neutron skin thickness @xmath166 of @xmath1pb is uniquely fixed by the slope parameter @xmath3 at @xmath102 @xmath5 @xcite , we can also extract a constraint @xmath167 fm for @xmath1pb by using the optimized parameters together with @xmath103 mev and @xmath6 mev .
our result is consistent with the estimated range @xmath168 fm in ref .
@xcite obtained by analyzing the experimental @xmath20 in @xmath1pb with an empirical range of @xmath169 .
one can see that our present constraint on @xmath166 of @xmath1pb has higher precision , indicating a more precise constraint on the symmetry energy at a subsaturation density is very helpful to extract @xmath166 of @xmath1pb from the electric dipole polarizability .
our result further agrees with the constraint @xmath170 fm obtained from the @xmath1pb dipole polarizability by using an empirical correlation between @xmath20 and @xmath166 of @xmath1pb @xcite , the constraint @xmath171 fm extracted very recently from coherent pion photoproduction cross sections @xcite , and within the experimental error bar the constraint @xmath172 fm extracted from the prex at jlab @xcite .
furthermore , it has been well established that the core - crust transition density @xmath173 in neutron stars , which plays a crucial role in neutron star properties @xcite , is strongly correlated with the density slope @xmath12 of the symmetry energy ( see , e.g. , ref .
in particular , in ref .
@xcite , the same correlation analysis method as in this work has been successfully applied to study the correlation between @xmath173 and the various macroscopic quantities , and indeed a strong correlation between @xmath173 and @xmath12 has been found . as mentioned in ref .
@xcite , a similar strong correlation is also existed between @xmath173 and @xmath3 , and this is demonstrated in fig .
[ rhot ] which shows the same correlations as fig .
[ alphad ] but for the core - crust transition density @xmath173 in neutron stars . here
, the transition density @xmath173 is calculated by using a dynamical approach ( see , e.g. , ref .
one can see from fig .
[ rhot ] that , for both @xmath174 and @xmath96 @xmath5 , @xmath173 exhibits a strong correlation with @xmath100 , a weak dependence on @xmath175 and @xmath176 , but almost no sensitivity to other macroscopic parameters . employing the optimized values for other macroscopic parameters as well as @xmath103 mev and @xmath6 mev , we then obtain a value of @xmath177 @xmath5 , which agrees well with the empirical values @xcite .
in summary , we have demonstrated that the electric dipole polarizability @xmath20 in @xmath1pb is sensitive to both the magnitude @xmath2 and density slope @xmath3 of the symmetry energy at a subsaturation cross density @xmath4 @xmath5 , and it decreases ( increases ) with @xmath2 ( @xmath3 ) , leading to a positive correlation between @xmath3 and @xmath2 for a fixed value of @xmath20 in @xmath1pb . using the experimental value of @xmath20 in @xmath1pb measured at rcnp and the very well - constrained range of @xmath2 ,
we have obtained a strong constraint on the slope parameter @xmath6 mev .
this constraint is in surprisingly good agreement with the previous solely existing constraint @xmath143 mev from neutron skin data of sn isotopes , demonstrating the robustness of these constraints on the value of the @xmath3 parameter .
the present constraint of @xmath3 further leads to @xmath145 mev and @xmath178 mev for the symmetry energy at saturation density , the neutron skin thickness @xmath179 fm for @xmath1pb , and @xmath180 @xmath5 for the core - crust transition density of neutron stars .
these results are nicely consistent with many other constraints extracted from terrestrial experiments , astrophysical observations , and theoretical calculations with controlled uncertainties .
our present results are based on the standard shf energy density functional .
it will be interesting to see how the results change if different energy - density functionals , e.g. , the relativistic mean field model or the extended non - standard shf energy density functional , are applied .
these works are in progress and will be reported elsewhere .
we are grateful to li - gang cao for helpful discussions on the skyrme - rpa code .
this work was supported in part by the major state basic research development program ( 973 program ) in china under contract nos .
2015cb856904 and 2013cb834405 , the nnsf of china under grant nos . 11135011 and 11275125 , the
shu guang " project supported by shanghai municipal education commission and shanghai education development foundation , the program for professor of special appointment ( eastern scholar ) at shanghai institutions of higher learning , and the science and technology commission of shanghai municipality ( 11dz2260700 ) . | nuclear structure observables usually most effectively probe the properties of nuclear matter at subsaturation densities rather than at saturation density . we demonstrate that the electric dipole polarizibility @xmath0 in @xmath1pb is sensitive to both the magnitude @xmath2 and density slope @xmath3 of the symmetry energy at the subsaturation cross density @xmath4 @xmath5 . using the experimental data of @xmath0 in @xmath1pb from rcnp and the recent accurate constraint of @xmath2 from the binding energy difference of heavy isotope pairs , we extract a value of @xmath6 mev .
the implication of the present constraint of @xmath3 to the symmetry energy at saturation density , the neutron skin thickness of @xmath1pb and the core - crust transition density in neutron stars is discussed . |
the study of the large scale structure of the universe is now entering a new phase . the two - degree field galaxy redshift survey ( hereafter 2dfgrs ;
colless et al .
2001 , 2003 ) and the sloan digital sky survey ( sdss ; york et al .
2000 ; adelman - mccarthy et al .
2006 ) have yielded high precision measurements of the power spectrum of galaxy clustering on large scales ( cole et al . 2005 ; tegmark et al .
2006 ; percival et al . 2007 ; padmanabhan et al .
when these measurements are combined with high angular resolution maps of the temperature fluctuations in the cosmic microwave background radiation ( e.g. hinshaw et al . 2003 , 2007 ) , tight constraints can be derived on many of the parameters in the cold dark matter model ( e.g. sanchez et al .
2006 ; spergel et al .
2007 ) . within this context
, the emphasis in large scale structure studies is shifting to measuring the clustering signal for samples of galaxies defined by intrinsic properties such as luminosity , colour , morphology or spectral type , with the goal of developing our understanding of the physics of galaxy formation .
the sdss and 2dfgrs catalogues contain sufficiently large numbers of galaxies over a large enough volume to allow robust measurements of clustering to be made for such subsamples ( e.g. norberg et al .
2001 , 2002a ; zehavi et al .
2002 , 2004 , 2005 ; madgwick et al .
2003 ) .
there are many observational clues which point to a dependence of galaxy properties on their local environment .
dressler ( 1980 ) argued that galaxies follow a morphology - density relation , with the fraction of early - type galaxies increasing with the local density .
galaxy clusters have well defined red sequences in the colour - magnitude relation ( bower , lucey & ellis 1992 ; stanford , eisenhardt & dickinson 1998 ) .
analyses of the 2dfgrs and sdss data sets have probed the connection between density and galaxy colour or type over a wider range of environments than was previously possible to reveal a general bimodality in the galaxy population ( e.g. lewis et al .
2002 ; hogg et al .
2003 , 2004 ; balogh et al .
these results suggest a tight correlation between the nature of a galaxy and its local environment , or equivalently the mass of its host dark matter halo .
measures of the clustering amplitude of different populations of galaxies reveal different results , indicating that these populations sample the underlying mass distribution in different ways .
hence , such clustering measurements can potentially tell us how the efficiency of the galaxy formation process depends upon halo mass .
the 2dfgrs and sdss allow us to push measurements of galaxy clustering beyond the traditional two - point correlation function or power spectrum . by extending the clustering analysis to higher orders
, we can extract new information about the connection between galaxies and matter . in previous work , we employed a counts - in - cells analysis to measure the higher - order correlation functions ( baugh et al .
2004 ; croton et al .
2004b ) and the void probability function ( croton et al .
2004a ) for galaxies samples of different luminosity extracted from the 2dfgrs .
we found that the higher - order correlation functions measured for 2dfgrs galaxies follow a hierarchical scaling pattern , characteristic of the clustering pattern which results from the growth of initially gaussian fluctuations due to gravitational instability .
croton et al .
( 2004a ) obtained independent confirmation of this result when they uncovered an unambiguous universal form for the void probability function .
the values of the correlation functions do , however , show differences from the expectation for the best fitting cold dark matter model .
gaztaaga et al .
( 2005 ) measured the three - point correlation function without averaging over a cell volume , using triangles of galaxies with sides of varying length ratios .
they found the first clear evidence for a non - zero second order or nonlinear bias parameter , suggesting that the relation of galaxies to the underlying dark matter may be more complicated than previous analyses had suggested ( e.g. verde et al .
2002 ; lahav et al .
measuring the higher - order correlation functions is a challenging task , even with surveys of the size of the 2dfgrs and the sdss .
baugh et al .
( 2004 ) found that their measurements of the correlation function for @xmath4 galaxies in the 2dfgrs were affected on large scales by the presence of associations of rich clusters ; nichol et al .
( 2006 ) found similar effects in the sdss . in this paper
, we extend our earlier work by exploiting the availability of colour information for the 2dfgrs catalogue . in section 2
we briefly describe the 2dfgrs galaxy catalogue , the counts - in - cells approach we use and the statistics we measure , along with a reprise of how higher - order clustering measurements can be used to make deductions about galaxy bias . in section 3
we present our results for the higher - order clustering of 2dfgrs galaxies selected by both luminosity and colour .
we give a simple interpretation of these results in section 4 and present a summary in section 5 .
the methodology we use is identical to that described by baugh et al .
( 2004 ) and croton et al .
( 2004a ; 2004b ) .
these authors measured the higher - order clustering and the void probability function for galaxies as a function of luminosity , analysing volume limited samples drawn from the 2dfgrs . in this paper
we extend this earlier work to consider samples defined by galaxy colour in addition to luminosity , as we did for the case of the 3-point correlation function in gaztaaga et al .
full details of the clustering measurements and a complete discussion of their interpretation can be found in the above references ; in this section , for completeness , we provide a brief outline of the 2dfgrs ( 2.1 ) , the statistics measured ( 2.2 and 2.3 ) and give a recap of the implications of the higher - order clustering statistics for galaxy bias ( 2.4 ) .
our analysis employs the completed 2dfgrs ( colless et al . 2001 ; 2003 ) which contains a total of 221,414 unique , high quality galaxy redshifts down to a nominal magnitude limit of @xmath5 with a median redshift @xmath6 .
in addition to @xmath7-band magnitudes , @xmath8-band images have now been scanned , allowing a @xmath9 colour to be defined for each galaxy . to maximise the volume sampled ,
we restrict ourselves to regions with spectroscopic completeness in excess of @xmath10 ; however , in practice , the typical completeness is much higher than this in the final 2dfgrs , with the mean completeness in spheres of radius @xmath11mpc better than @xmath12 .
the remaining incompleteness is accounted for using the volume weighting corrections described in croton et al .
( 2004b ) .
in order to construct a volume limited subsample from a flux limited redshift survey , it is necessary to model the redshift dependence of a galaxy s luminosity or magnitude in the passband in which the survey selection is defined .
we apply the colour dependent @xmath13-correction model of cole et al .
( 2005 ) to account for the average change in galaxy magnitude due to redshifting of the @xmath7-filter bandpass ( the `` k - correction '' ) and also the associated typical galaxy evolution ( the `` e - correction '' ) .
the purpose of defining a volume limited sample is to isolate galaxies of similar intrinsic luminosity in a sample with a simple radial selection function to facilitate clustering measurements ( see norberg et al . 2001 ; 2002a ) .
once such a sample has been constructed , we can subdivide it by galaxy colour , using the @xmath7 and @xmath8-band photometry available for 2dfgrs data .
cole et al .
( 2005 ) analysed the rest frame @xmath14 colour distribution of galaxies in the 2dfgrs and found a clear division at a rest frame colour of @xmath15 .
the bimodality in the colour distribution about this reference point leads to natural definition of `` red '' and `` blue '' galaxies .
this separation by colour is similar , though by no means identical to the selection by spectral type employed in previous 2dfgrs analyses ( norberg et al .
2002a ; madgwick et al .
2002 , 2003 ; conway et al . 2005 ; croton et al . 2005 ; see figure 2 of wild et al .
2005 ) . the volume limited samples used in this paper are listed in table 1 ( see also table 1 of croton et al .
2004b for further properties of these samples .
note that the improved @xmath13-correction used here has resulted in small differences to the total number of galaxies falling within each volume limited boundary when compared with our previous work ) .
our analysis covers a wide range of luminosity , from @xmath16 0.2@xmath17to @xmath16 4 @xmath17 ( the mean effective luminosities of the volume limited samples differ by a factor of 6 from the faintest to brightest ) .
the error bars plotted on our measurements in each of the figures are jackknife estimates derived by subdividing each sample into 20 areas on the sky .
our measurements are strongly correlated from bin to bin , so it is essential to take this into account when fitting models to the measurements . for the purpose of deriving confidence intervals on fitted parameters , we compute a full covariance matrix using the ensemble of mock 2dfgrs catalogues described by norberg et al .
( 2002b ) .
see section 4 for further details on how errors from the mocks are used in our analysis .
the clustering statistics we employ require an accurate measurement of the count probability distribution function ( cpdf ) for the red and blue galaxy populations in our volume limited samples .
this is done using a counts - in - cells ( cic ) analysis .
the cpdf for a given smoothing scale is measured by throwing down a large number ( @xmath18 ) of spheres of radius @xmath19 within the survey volume .
the probability of finding exactly @xmath20 galaxies within a sphere of this scale is given by : @xmath21 here @xmath22 is the cpdf at the given scale @xmath19 , where @xmath23 is the number of spheres containing @xmath20 galaxies out of a total of @xmath24 spheres used . from the cpdf
, the clustering moments of the distribution can be calculated directly .
for example , the volume - averaged 2-point correlation function is given by @xmath25 - 1/\bar{n } , \ ] ] where the mean number of galaxies expected in the sphere , @xmath26 , is simply @xmath27 expressions for the higher - order moments in terms of the cpdf are given in appendix a of gaztaaga ( 1994 ) .
in the hierarchical model of galaxy clustering , all higher - order correlations can be expressed in terms of the 2-point function , @xmath28 , and dimensionless scaling coefficients , @xmath29 ( see bernardeau et al .
2002 ) : @xmath30 traditionally , @xmath31 is referred to as the _ skewness _ of the distribution and @xmath32 as the _ kurtosis_. given that both @xmath28 and @xmath33 ( @xmath34 ) can be evaluated using the cpdf , the hierarchical amplitudes @xmath35 can readily be measured and the scaling behaviour of eq .
[ eq : sp ] tested ( i.e. the validity of the @xmath36 power law dependence for higher - order correlations ; see baugh et al . 2004 ) .
ccccccccc population & & @xmath37 & @xmath38 & @xmath39 & @xmath40 & @xmath41 & @xmath42 + & & & ( eq . [ eq : br ] ) & ( eq . [ eq : c2 ] ) & & & + + red galaxies & -18.0 & -19.0 & 7710 & @xmath43 & @xmath44 & @xmath45 & @xmath46 & @xmath47 + & -19.0 & -20.0 & 18693 & @xmath48 & @xmath49 & @xmath50 & @xmath51 & @xmath52 + & -20.0 & -21.0 & 15147 & 1 & 0 & @xmath53 & @xmath54 & @xmath55 + + blue galaxies & -18.0 & -19.0 & 14086 & @xmath56 & @xmath57 & @xmath58 & @xmath59 & @xmath60 + & -19.0 & -20.0 & 22499 & @xmath61 & @xmath62 & @xmath63 & @xmath64 & @xmath65 + & -20.0 & -21.0 & 15125 & 1 & 0 & @xmath66 & @xmath67 & @xmath68 + + red+blue galaxies & -18.0 & -19.0 & 21796 & @xmath69 & @xmath70 & @xmath71 & @xmath72 & @xmath73 + & -19.0 & -20.0 & 41192 & @xmath74 & @xmath75 & @xmath76 & @xmath77 & @xmath78 + & -20.0 & -21.0 & 30272 & 1 & 0 & @xmath79 & @xmath80 & @xmath81 + [ table1 ] a complementary way in which to study the hierarchical clustering paradigm is through the reduced void probability function ( vpf ) .
put simply , the reduced vpf , @xmath82 , is a parametrisation of @xmath83 , the probability of finding an empty sphere in the galaxy distribution , in terms of the expectation from a purely poisson galaxy distribution : @xmath84 white ( 1979 ) derived eq .
[ eq : chi ] after writing the void probability function as a power series expansion in the moments of the cpdf ( @xmath0 ) to _ all _ orders . under the hierarchical ansatz of eq .
[ eq : sp ] , this expansion can be expressed as a function of @xmath85 only : @xmath86 therefore , if the hierarchical assumption is valid , a plot of @xmath82 as a function of @xmath85 should produce a universal curve for galaxy catalogues with different mean densities and clustering properties , assuming common hierarchical amplitudes for the different populations ( we will test this assumption in section [ sec : sp ] ) .
note that clustered tracers which deviate strongly from a poisson distribution will have vpf values @xmath87 .
the precise form of the reduced vpf is set by the hierarchical amplitudes , @xmath29 , for which different models of clustering predict different values .
here we summarise only the most successful model , the _ negative binomial model _ , which does very well at reproducing the measurements from the 2dfgrs for both the vpf ( fig . 3 in croton et al .
2004b ) and the hierarchical amplitudes , @xmath29 ( table 1 in baugh et al .
the reduced vpf and hierarchical amplitudes for the negative binomial model are @xmath88 further analytic models , like the minimal and thermodynamic , are explored in croton et al .
. the distribution of galaxies could be quite different from that of the underlying dark matter .
one could conceive of physical processes which could lead to a dependence of the efficiency of galaxy formation on the mass and perhaps the environment of dark matter hales ( e.g. croton , gao & white 2007 ) .
if this is indeed the case , galaxies should be regarded as _ biased _ tracers of the matter distribution .
a simple model for galaxy bias was introduced by fry & gaztaaga ( 1993 ) .
these authors assumed that the density contrast in the galaxy distribution , @xmath89 , can be expressed as a general non - linear function of the density contrast of the dark matter , @xmath90 , so that : @xmath91 $ ] . for density fluctuations smoothed on large enough scales
so that the matter density contrast is of the order unity or smaller , this relation can be expanded in a taylor series : @xmath92 on scales where the variance , @xmath93 , is small , the leading order contribution to the variance is dominated by the linear term : @xmath94 ^ 2 , \label{eq : b1}\ ] ] where @xmath95 is the so - called `` linear bias '' , @xmath96 .
the leading order term in the expansion for the skewness , @xmath40 , is : @xmath97 where we use the notation @xmath98 ( expressions for the hierarchical amplitudes up to @xmath99 are given in fry & gaztaaga 1993 ) .
@xmath100 encodes the non - linear gravitational evolution of the matter distribution from gaussian initial conditions .
hence , two non - linear effects are present in @xmath101 : gravity ( i.e. as contained in @xmath100 ) and galaxy bias ( i.e. as quantified by @xmath102 )
. both terms are modulated by the linear bias , @xmath95 .
therefore , in general , the interpretation of the skewness measured in the galaxy distribution is not trivial .
however , on weakly non - linear scales we have detailed models of what to expect for the skewness of the dark matter ( see , for example , bernardeau et al .
2002 ; calculations have also been carried out for non - standard gravitational models and cosmologies see gaztaaga & lobo 2001 ) .
we can therefore hope to learn about non - linear galaxy biasing if we can measure the skewness in the galaxy distribution , @xmath103 , on scales on which the underlying fluctuations are only weakly nonlinear .
it is useful to define a _ relative _ bias to aid with the interpretation of measurements of higher - order clustering .
the relative bias describes the change in clustering signal compared to that measured for a reference sample . using eq .
[ eq : b1 ] as a guide , we define the relative linear bias of a sample , @xmath104 , as the square root of the ratio of the @xmath105-point correlation function ( or variance ) measured for the sample , relative to that found for the reference sample , denoted by an asterisk : @xmath106 our reference sample is the volume limited population of galaxies in the magnitude range @xmath107 , which is one magnitude brighter than the one used in croton et al .
( 2004a ) .
the reason for choosing a brighter sample is directly related to the influence the large coherent superstructures have on the @xmath4 sample : it makes more sense to use as reference a sample which is not systematically affected by the presence of such rare massive structure . for completeness , in table 1 we show the relative bias results for the volume limited `` all colour '' catalogues analysed in croton et al .
( 2004a ) , however now presented relative to this brighter reference sample . on scales
for which a linear bias is a good approximation , i.e. when @xmath108 for @xmath109 , we can relate the @xmath110 measured for different galaxy samples regardless of the value of @xmath111 for the underlying mass : @xmath112 more generally , if galaxy bias is non - linear one can introduce a measure to quantify the relative non - linear bias : @xmath113 where an asterisk denotes a quantity measured for the reference sample . in general , if the reference sample is un - biased ( i.e. @xmath114 and @xmath115 ) , we then have @xmath116 and @xmath117 for all samples .
in this section we present the main results of the paper , which are inferred from our measurement of the count probability distribution function ( cpdf ) , as outlined in section 2 , for the samples of red and blue galaxies listed in table 1 .
we present the volume averaged correlation functions , @xmath0 , in 3.1 , the hierarchical amplitudes , @xmath29 , in 3.2 , and the reduced void probability function ( vpf ) in 3.3 .
[ fig1 ] shows the volume averaged correlation functions for orders @xmath118@xmath119 as a function of smoothing scale .
the left hand panel shows the measurements for red galaxies and the right panel for blue galaxies . in both panels
the symbols show the results for samples in the magnitude range @xmath120 ( i.e. approximately @xmath121 ) , while the adjacent dashed and dotted lines show samples which are one magnitude brighter and fainter respectively .
the order of volume averaged correlation function is indicated by the symbol type shown in the legend .
red galaxies clearly show a larger clustering amplitude than the equivalent population of blue galaxies , as expected from earlier results obtained for the two - point correlation function as a function of colour ( e.g. zehavi et al .
2005 in the sdss ) or spectral type ( norberg et al . 2002a
; madgwick et al .
2003 in the 2dfgrs ) .
the dependence of clustering strength on galaxy luminosity appears to be the strongest for the blue population , with the faintest blue galaxies being more weakly clustered than the brightest .
the situation is more complicated for red galaxies ; both the faintest and brightest samples of red galaxies appear to be more strongly clustered than the red @xmath122 sample .
there is even a suggestion that the faintest red galaxies become the more strongly clustered with increasing order , although the errors on the measurements are also increasing at a fixed scale for higher values of @xmath123 .
this type of behaviour was noted previously for the two - point correlation function .
norberg et al .
( 2002b ) found an increase in the correlation length of early spectral types for samples faintwards of @xmath4 , with the faintest early types displaying a similar correlation length to the brightest early types .
zehavi et al .
( 2005 ) found a similar increase in clustering strength for faint red galaxies .
this trend can be readily explained ; the satellite population in massive clusters is predominantly made up of faint , red galaxies and clusters are known to be strongly biased tracers of the mass distribution ( e.g. padilla et al . 2004 ) .
[ fig2 ] extends the results presented in the previous subsection by plotting the hierarchical amplitudes ( as defined by eq .
[ eq : sp ] ) , @xmath40 ( the skewness ) , @xmath41 ( the kurtosis ) and @xmath42 , as a function of scale .
this is done for red galaxies in the left panel and blue galaxies in the right panel .
again , the results are shown for samples with different luminosities , with symbols representing the measurement of galaxies in the magnitude range @xmath120 , and dashed and dotted lines showing the @xmath29 for samples one magnitude brighter and fainter than this respectively .
red galaxies show a strong segregation of the hierarchical amplitudes with luminosity at a given clustering order ; the @xmath29 for blue galaxies show little dependence on luminosity . on large scales , n - body simulations and
perturbation theory suggest that the hierarchical amplitudes for the mass should only vary slowly with scale in the case of initially gaussian fluctuations which grow through gravitational instability ( e.g. juskiewicz , bouchet & colombi 1993 ; baugh et al .
1995 ; for a summary of the theoretical predictions , see bernardeau et al .
observationally , the hierarchical amplitudes for galaxies have traditionally been measured to be either approximately constant or decreasing slowly out to progressively larger smoothing scales ( e.g. gaztaaga 1992 ; bouchet et al .
1993 ; gaztaaga 1994 ; hoyle , szapudi & baugh 2000 ; frith , outram & shanks 2006 ; see bernardeau et al .
2002 for a review of previous measurements ) .
however , recent work has shown that the presence of large , rare superstructures in the galaxy distribution can produce a significant upturn in the value of @xmath29 on large scales ( e.g. szapudi & gaztaaga 1998 ; baugh et al . 2004 ; croton et al .
2004b ; nichol et al .
according to figure 10 from gaztaaga et al .
( 2005 ) , the upturn in the hierarchical amplitudes appears to be equally important when using measurements of the 3-point correlations function of red and blue galaxies from the 2dfgrs @xmath124 sample .
nevertheless , it should be emphasised that , in the case of the 2dfgrs , the impact of the superstructures on the measured hierarchical amplitudes is most pronounced in the @xmath125 volume limited sample and is essentially negligible in the brighter and fainter samples .
this is due to the superstructures falling entirely within the redshift interval spanned by the @xmath4 sample . in the case of the fainter volume limited sample ,
the superstructures lie mostly beyond the maximum redshift which defines the sample , whilst in the case of the brighter sample , the volume covered is much larger and the significance of the superstructures is correspondingly lower .
see croton et al .
( 2004b ) , baugh et al .
( 2004 ) and gaztaaga et al .
( 2005 ) for detailed investigations into how these two superstructures affect the measurement of higher - order statistics .
the flatness of the earlier results for the @xmath29 can be understood by noting that a ) most smaller surveys never intersected with these rare superstructures of galaxies , b ) their influence for larger surveys , such as the original 2dfgrs parent catalogue , the apm survey ( i.e. gaztaaga 1994 ) or 2mass ( frith , outram & shanks 2006 ) , was diluted by the significant increase in sample volume , and c ) typically any contribution from large coherent superstructures is downgraded in a flux limited catalogue rather than a volume limited sample due to the varying selection function . in fig .
[ fig2 ] , the grey shaded regions show the results for the hierarchical amplitudes measured from the @xmath121 sample after removing the two large superstructures identified in baugh et al .
the shaded regions indicate the extent of the @xmath126 errors on the measurements and should be compared with the symbols and error bars , which show the results for the same volume limited sample but _ without _ removing the superstructures from the cic analysis . by omitting the two superstructures ,
the theoretically expected constant behaviour of @xmath127 is restored for larger smoothing lengths .
however , whilst the results change on large scales for @xmath41 and @xmath42 on removing the superstructures , there is still significant scale dependence of the @xmath29 with opposite gradients found for red and blue galaxies . from fig .
[ fig2 ] , the impression is that the red population is the most strongly affected by the presence of the superstructures can be derived . ] .
this behaviour should be contrasted with the measurements by croton et al .
( 2004b ) and baugh et al .
( 2004 ) made using the full l@xmath128 sample . in these papers , removing the superstructures from the cic analysis , led , to within the accuracy of the measurements , to all the hierarchical moments up to 5@xmath129 order becoming roughly independent of scale .
our results for red and blue galaxies imply that the @xmath122 results for all galaxies were due to a fortuitous cancellation of the trends with scale seen for red and blue galaxies .
finally , it is worth noting that the upturn seen on small scales in both @xmath41 and @xmath42 for the bright and faint populations of blue galaxies is not statistically significant : in both cases , the jackknife errors , if plotted , are fully consistent with no upturn at all . in fig .
[ fig3 ] we plot the reduced void probability function ( vpf ) for the three luminosity samples ( using different symbols as shown by the legend ) and for red and blue galaxy populations ( left and right hand panels respectively ) .
previous measurements have revealed that galaxy samples defined by luminosity ( or equivalently number density ) display a universal form for the vpf ( e.g. croton et al .
2004a ; hoyle & vogeley 2004 ; patiri et al .
this universality also holds for samples at different redshifts ( conroy et al . 2005 ; 2007 ) .
the recovery of a universal form for the vpf is evidence for the hierarchical scaling of the higher - order correlation functions of galaxies ( e.g. eq [ eq : sp ] ) .
the actual form of the vpf curve is well described , in redshift space , by the negative binomial model ( see also fry 1986 and gaztaaga & yokoyama 1993 ) . here
we report a significant departure from the universal scaling of the vpf . as can be seen in fig .
[ fig3 ] , the vpf measured for _ red _ galaxies deviates from the universal scaling in a way such that the deviation is strongest for the faintest galaxies considered .
the vpf measured for blue galaxies , on the other hand , shows no dependence on luminosity , and matches almost perfectly the negative binomial model across the entire range plotted . note that the deviation seen for the red @xmath121 population appears to arise from the contribution of the two superstructures discussed in the previous subsection .
once these large structures are removed from the analysis and the vpf is recalculated , the results for the @xmath121 sample shift back to the negative binomial model , as shown by the grey shaded region . on the other hand , the fainter red galaxy sample ( shown by the triangles in fig .
[ fig3 ] ) , which is not affected by the two superstructures , does not agree with the negative binomial model .
this disagreement persists if we consider even fainter samples , which we have not plotted in the figure for clarity .
our results show that once we have accounted for the difference in mean density and clustering , faint red galaxies have a characteristically larger vpf than is measured for the other galaxy samples . from eq .
[ eq : vpf ] , we expect that samples with a larger vpf should also also have larger values for the @xmath29 , and this is borne out by the results given in table 1 , which are discussed in the next section .
a high value for @xmath29 indicates non - linear biasing . if one were to simply increase the _ linear _ bias , this would increase the amplitude of the correlation functions , @xmath0 , but would reduce the hierarchical amplitudes ( see eq . [ eq : sp ] and eq .
[ eq : sjlinear ] ) . a non - zero non - linear bias is characteristic of the high density regions of the dark matter distribution .
this is in line with the suggestion made above that the red faint galaxies could be made up of a significant population of satellites in clusters .
the reasons behind the success of the negative binomial model for the vpf are unclear .
vogeley et al .
( 1994 ) showed that the vpf measured in their simulations only agreed with the negative binomial model when the clustering pattern was distorted by peculiar motions ; the vpf measured in real space did not match the negative binomial model .
however , our results imply that the peculiar motions of galaxies are not the primary agent responsible for the success of the negative binomial model .
we find that the faint red galaxies , which are primarily satellites within galaxy clusters and hence likely to display large peculiar motions , in fact show the strongest departure from the vpf predicted by the negative binomial model , rather than being driven towards this model .
in this section we give some simple interpretations of the measurements of the hierarchical amplitudes reported in section [ sec : sp ] .
this requires us to perform fits to the measurements , in the first case by extracting best fitting values for the hierarchical amplitudes , @xmath29 ( as defined in eq .
[ eq : sp ] ) , and then to constrain the linear and non - linear bias parameters ( see eq .
[ eq : br ] and eq .
[ eq : c2 ] ) . measures of the hierarchical amplitudes on different smoothing scales are correlated , a fact which has often been ignored in previous analyses reported in the literature . the existence of a bin - to - bin correlation is clearly demonstrated by the response of our estimates of the hierarchical amplitudes to the removal of the two largest superstructures in the 2dfgrs . given this correlation , it is necessary to construct a covariance matrix for the measurements in order to carry out meaningful fits .
one way to do this is to use a jackknife approach in which the survey volume is split into a number of equal sized subvolumes , on the order of 20 ( if not more to construct a stable covariance matrix ) .
unfortunately , this is infeasible for our fainter volume limited samples , as the volume covered in these cases is too small to be subdivided an appropriate number of times .
instead , we use the ensemble of mock 2dfgrs catalogues whose construction is described in norberg et al .
( 2002b ) .
particles were chosen from an n - body simulation to represent galaxies using a simple parametric function of the local smoothed density of dark matter ( see cole et al .
1998 for a description of the bias algorithm ) .
the parameters in this empirical biasing prescription were constrained to reproduce the typical clustering measured in the full flux limited 2dfgrs ( hawkins et al .
consequently , one limitation of the mock catalogues is that the luminosity and colour assigned to a `` galaxy '' is independent of the local density of the particle .
this means that the mocks do not display any dependence of clustering on luminosity or colour .
also , no information about the higher - order clustering of mock galaxies was used to select the parameters in the bias prescription .
hence , although the two point correlation function for mock galaxies agrees very closely with that measured for the full 2dfgrs , there is no guarantee that the higher - order clustering in the mocks will look like that measured in the real survey .
as we found in our earlier work using the mocks ( croton et al .
2004b ; gaztaaga et al .
2005 ) , the covariance matrices estimated from the mocks are stable and give robust estimates of best fitting quantities and the associated errors .
a principal component decomposition of the covariance matrix reveals that the first few eigenvectors are typically responsible for the bulk of the variance or signal .
we rank the eigenvectors in order of decreasing variance and retain sufficient eigenvectors to account for , at least , 95% of the variance .
we first perform fits to extract values for the scale _ independent _ hierarchical amplitudes .
although our results may appear in some cases to show a strong dependence on cell radius ( fig .
[ fig2 ] ) , it is important to bear in mind that we can only show the diagonal part of the covariance matrix in plots and our fitting procedure takes into account the covariance between bins .
moreover , since there is a tradition of fitting constants to estimates of the @xmath29 over a range of scales , it is useful to repeat such an analysis for our measurements in order to facilitate comparisons with previous work . we restrict our attention to the @xmath29 values obtained for spheres with radii in the range @xmath130mpc .
these are slightly larger scales than we used in croton et al .
( 2004b ) , since we aim to avoid scales which may be affected by shot noise , and our samples , as a result of being split by colour , typically have half the number of galaxies in each luminosity bin than before . to help with a comparison to our earlier work we also re - analyse
the `` all colour '' galaxy catalogues across this new fitting range ( note that these galaxies also differ in their improved @xmath13 corrections , as described in section [ survey ] ) .
the results of a one parameter fit ( @xmath29 ) to the hierarchical amplitudes are quoted as a function of luminosity for all galaxy colours and also the red and blue populations separately in table [ table1 ] , in which we also give the 95 % confidence interval around the best fitting value .
next we constrain the relative bias parameters defined in eq .
[ eq : br ] and eq .
[ eq : c2 ] . recall that in this paper we define our reference sample to be volume limited galaxies in the magnitude range @xmath107 .
the error in the relative linear bias is straightforward to find using the mocks and the results are given in table 1 . obtaining a robust fit for the nonlinear bias , @xmath131 ,
is more subtle . by construction
, the mocks will return a zero nonlinear bias . ] , whereas the expectation value for 2dfgrs samples can be anything .
therefore , in order to get a reliable estimate of the error on the best fitting nonlinear bias using the mocks , we need to take an indirect approach .
we fit instead the combination @xmath132 , for which the mock catalogues are in rather good agreement with the real survey . taking the best fitting value for the quantity @xmath133 , and the skewness measured for the reference sample , we can obtain a best fitting value for the nonlinear bias , @xmath131 , using eq .
[ eq : c2 ] . the error on the best fitting value
is obtained by adding the errors on @xmath132 and the skewness of the reference sample in quadrature , ie .
assuming @xmath133 and @xmath134 are uncorrelated .
the resulting values of the nonlinear bias and the error on the best fit are listed in table 1 .
the results in table 1 reveal substantial differences in the hierarchical amplitudes obtained for red and blue galaxies and in the way in which these amplitudes change with luminosity .
overall , we find that blue galaxies display smaller hierarchical amplitudes than red galaxies .
red galaxies show a significant change in @xmath29 with luminosity , whereas blue galaxies show no such trend . assuming gaussian errors
, we find a 3-@xmath135 shift in the best fitting @xmath40 between the faintest and brightest samples of red galaxies . in view of the relatively small baseline in luminosity
over which we can perform such a fit , a factor of six in median luminosity moving from the faintest to the brightest volume limited sample , it is remarkable to see such a clear change in the higher - order clustering .
these results are in good agreement with those found in croton et al .
( 2004b ) ( see their fig .
it is also noteworthy that the relative bias @xmath136 quoted in table 1 , shows the opposite tendency to that displayed by the hierarchical amplitudes : there is little dependence of the linear relative bias on luminosity for red galaxies and a clear trend for blue galaxies .
finally , it is encouraging to note that the conclusions we draw from our counts - in - cells analysis are consistent with those we reported in gaztaaga et al .
( 2005 , sags - ii ) , who measured the 3-point correlation function for different triangle shapes and sizes ( i.e. without averaging over the volume of a cell ) .
in particular fig.7 in sags - ii shows how for both equilateral and elongated triangles , late galaxies show little luminosity segregation in @xmath137 , while early galaxies show a strong segregation , with @xmath137 increasing with decreasing luminosity ( see also nishimichi et al .
this paper complements and extends previous clustering analyses of the 2dfgrs , taking advantage of the availability of @xmath8-band photometry to study the spatial distribution of galaxies as a function of _ both _ their colour and luminosity .
we use a counts - in - cells approach to estimate the count probability distribution function ( cpdf ) of red and blue galaxies in different luminosity bins , spanning a factor of six in luminosity around @xmath122 . from the cpdf
, we estimate the volume - averaged higher - order correlation functions , @xmath0 , the hierarchical amplitudes , @xmath29 , and the reduced void probability function for our galaxy samples , @xmath82 .
we use our measurements of the hierarchical amplitudes to constrain the linear and nonlinear bias parameters in a simple model of galaxy clustering .
the main results of this paper can be summarised as follows : * _ colour segregation in @xmath0 : _ blue galaxies show significantly lower clustering amplitudes than red galaxies .
this trend holds up to the five point correlation function , extending the results previously reported for two - point correlations ( e.g. norberg et al .
2002a ; madgwick et al .
2003 ; zehavi et al .
* _ luminosity segregation in @xmath0 : _ blue galaxies display a monotonic increase of clustering strength with luminosity , mirroring the results found for the two - point function of 2dfgrs galaxies ( norberg et al .
the behaviour of the red population is more complicated .
faint and bright red galaxies are more clustered than @xmath122 red galaxies .
a hint of this behaviour was previously found for the two - point correlation function of early spectral types in the 2dfgrs ( which loosely correspond to the red sample in this paper ) by norberg et al .
( 2002a ) . *
_ luminosity segregation in @xmath29 : _ we fit scale - independent models to the hierarchical amplitudes which take into account the full covariance between adjacent bins . from this
we conclude that the hierarchical amplitudes of blue galaxies show little or no dependence on galaxy luminosity . on the other hand ,
the hierarchical amplitudes of red galaxies vary strongly with luminosity .
these new results , split by colour , explain why croton et al .
( 2004b ) found only a weak dependence of the @xmath29 on luminosity . *
_ superstructures in the 2dfgrs .
_ the hierarchical amplitudes for the red galaxies seems to be more affected by the presence of the superstructures . after removing them
, there is still significant scale dependence of the @xmath29 , with opposite gradients found for red and blue galaxies .
these results imply that the flatness of the @xmath122 @xmath29 measurements of croton et al .
( 2004b ) and baugh et al .
( 2004 ) , after removing the superstructures , are due to a fortuitous cancellation of the different trends with scale for red and blue galaxies .
* _ the reduced void probability function .
_ croton et al .
( 2004a ) found that the reduced void probability function measured for 2dfgrs samples defined by luminosity displayed a universal form , and matches the one predicted by a negative binomial distribution .
splitting the samples by colour , we find that blue galaxies show the universal reduced void probability function consistent with the negative binomial , but red galaxies do not .
the deviation from the negative binomial model is largest for faint red galaxies .
this result is seemingly at odds with previous interpretations of the success of the negative binomial model , in which it was suggested that galaxy peculiar motions were the primary agent behind the form of the reduced void probability function . *
_ linear and non - linear bias .
_ fry & gaztaaga ( 1993 ) introduced a simple model of galaxy bias in which the density contrast in galaxies is written as a taylor expansion of the density contrast in the underlying mass .
the first order bias term in this expansion is the common linear bias and the second order term is called the quadratic or nonlinear bias .
we use our measurements of the variance and the skewness ( @xmath29 ) to extract the linear and nonlinear bias parameters relative to the clustering in a reference sample .
the reference sample is the one for which we are able to make our best measurements of galaxy clustering . in this paper , galaxies with magnitudes in the range @xmath138
are treated as the reference sample ; there is a reference sample for red galaxies and one for blue galaxies .
the relative bias parameters extracted for red and blue galaxies are different .
the faintest red galaxies we consider have a linear bias consistent with that of the reference sample ; red @xmath122 galaxies have a linear bias below unity . for blue galaxies , the linear bias increases with luminosity .
the relative non - linear bias for red galaxies is positive whereas that extracted for blue galaxies is negative . in all cases , these offsets are significant at the @xmath139 level . our measurements of the higher - order clustering of galaxies as a function of colour and luminosity provide valuable new constraints on models of galaxy formation . while the colour of a galaxy is determined by its star formation history , the clustering of galaxies , at least on the scales probed in this paper , is driven primarily by the mass of the host dark matter halo and to a lesser extent by the formation history of the halo ( gao et al . 2005 ; harker et al .
2006 ; wechsler et al . 2006 ; croton , gao & white 2007 ) . semi - analytical models of galaxy formation make _ ab initio
_ predictions for the star formation histories of galaxies in a cosmological setting ( see the review by baugh 2006 ) .
this is done through simple physical prescriptions that describe those aspects of galaxy evolution believed to be important , including the rate of gas cooling , the timescale for star formation in galactic disks , mergers between galaxies , and feedback processes such as heating by supernova explosions or the accretion of material onto supermassive black holes .
these processes are poorly understood and consequently , there is no unique way in which to model them .
at present there are a number of galaxy formation models which follow the evolution of disks and spheroids from high redshift to the present day ( e.g. baugh et al . 2005 ; croton et al . 2006 ; bower et al . 2006 ; cattaneo et al . 2006
; de lucia & blaizot 2007 ) .
the parameters that constrain the simple physics assumed in these models are typically set without reference to galaxy clustering .
higher - order clustering and non - linear bias measures are sensitive to how galaxies populate groups and clusters , and their periphery .
we therefore look towards such statistics to further constrain and discriminate between different possible ( and plausible ) implementations of the galaxy formation physics .
the higher - order spatial distribution of galaxies provides an additional window through which we can hope to understand the complex physics governing galaxy evolution .
the 2dfgrs was undertaken using the two - degree field spectrograph on the anglo - australian telescope .
we acknowledge the efforts of all those responsible for the smooth running of this facility during the course of the survey , the 2dfgrs team for carrying out the observations and releasing the data and also the support of the time allocation committees during the project . this work was supported in part by nsf grants
ast00 - 71048 and ast05 - 07428 , the royal society through the award of a joint project grant , and by the european commission through the alfa - ii programme s funding of the latin american european network for astrophysics and cosmology ( lenac ) .
dc is supported by a deep2 postdoctoral fellowship .
pn acknowledges receipt of a pparc pdra fellowship held at the ifa .
eg acknowledges the spanish ministerio de ciencia y tecnologia ( mec ) , project aya2006 - 06341 with ec - feder funding , research project 2005sgr00728 from generalitat de catalunya , the galileo galilei institute for theoretical physics for hospitality , and the infn for partial support during the completion of this work .
cmb is funded by a royal society university research fellowship .
adelman - mccarthy j.k . , 2006 , apjs , 162 , 38 balogh m. , et al . , 2004 , mnras , 348 , 1355 baugh c.m . ,
gaztaaga e. , efstathiou , g. , 1995 , mnras , 274 , 1049 baugh c.m .
, et al . , 2004 , mnras , 351 , l44 baugh c.m .
, et al . , 2005 , mnras , 356 , 1191 baugh , c. m. 2006 , reports of progress in physics , 69 , 3101 bernardeau f. , colombi s. , gaztaaga e. , scoccimarro r. , 2002 , phys . rep . , 367 , 1 bouchet f.r . ,
strauss m.a . , davis m. , fisher k.b . , yahil a. , huchra j.p . , 1993 ,
apj , 417 , 36 bower r.g . , lucey j.r . ,
ellis , r.s .
, 1992 , mnras , 254 , 601 bower , r. g. , benson , a. j. , malbon , r. , helly , j. c. , frenk , c. s. , baugh , c. m. , cole , s. , & lacey , c. g. 2006 , mnras , 370 , 645 cattaneo a. , dekel a. , devriendt j. , guiderdoni b. , blaizot j. , 2006 , mnras , 370 , 1651 cole s. , hatton s. , weinberg d.h .
, frenk c.s . , 1998 , mnras , 300 , 656 cole s. , et al .
, 2005 , mnras , 362 , 505 colless m.m . , et al . , 2001 ,
mnras , 328 , 1039 colless m.m . , et al . , 2003 , astro - ph/0306581 conroy c. , et al . , 2005 , apj 635 , 990 conroy , c. , et al .
2007 , apj , 654 , 153 conway e. , et al . , 2005 , mnras , 356 , 456 croton d.j .
, et al . , 2004a , mnras 352 , 828 croton d.j . , et al . , 2004b , mnras 352 , 1232 croton , d. j. , et al .
2005 , mnras , 356 , 1155 croton d.j . , et al . , 2006 ,
mnras 365 , 11 croton , d. j. , gao , l. , & white , s. d. m. 2007 , mnras , 374 , 1303 de lucia , g. , & blaizot , j. 2007 , mnras , 375 , 2 dressler a. , 1980 , apj , 236 , 351 frith , w. j. , outram , p. j. , & shanks , t. 2006 , mnras , 373 , 759 fry j.n . , 1986 , apj , 308 , l71 fry j.n . , gaztaaga e. , 1993 , apj , 413 , 447 gao , l. , springel , v. , & white , s. d. m. 2005 , mnras , 363 , l66 gaztaaga e. , 1992 , apj , 398 ,
l17 gaztaaga e. , 1994 , mnras , 268 , 913 gaztaaga e. , yokoyama j. , 1993 , apj , 403 , 450 gaztaaga e. , lobo j.a . , 2001 , apj , 548 , 47 gaztaaga , e. , norberg , p. , baugh , c. m. , & croton , d. j. 2005 , mnras , 364 , 620 harker g. , cole s. , helly j. , frenk c. , jenkins a. , 2006 , mnras , 367 , 1039 hawkins e. , et al . , 2003 , mnras , 346 , 78 hinshaw g. , et al .
, 2003 , apjs , 148 , 119 hinshaw , g. , et al .
2007 , apj , 170 , 288 hogg d.w .
, et al . , 2003 , apj , 585 , l5 hogg d.w .
, et al . , 2004 , apj , 601 , l29 hoyle f. , & vogeley m.s . , 2004 ,
apj 607 , 751 hoyle f. , szapudi i. , baugh c.m . , 2000 , mnras , 317 , l51 juszkiewicz r. , bouchet f.r . , colombi s. , 1993 , apj , 412 , l9 lahav o. , et al .
, 2002 , mnras , 333 , 961 lewis i. , et al . , 2002 , mnras , 334 , 673 madgwick d.s .
, et al . , 2002 , mnras , 333 , 133 madgwick d.s .
, et al . , 2003 , mnras , 344 , 847 nichol r.c . , et al . , 2006 , mnras , 368 , 1507 nishimichi , t. , kayo , i. , hikage , c. , yahata , k. , taruya , a. , jing , y. p. , sheth , r. k. , & suto , y. 2007 , pasj , 59 , 93 norberg p. , et al
, 2001 , mnras , 328 , 64 norberg p. , et al . , 2002a , mnras , 332 , 827 norberg p. , et al .
, 2002b , mnras , 336 , 907 padilla n.d .
, et al . , 2004 ,
mnras , 352 , 211 padmanabhan , n. , et al .
2007 , mnras , 378 , 852 patiri , s. g. , betancort - rijo , j. e. , prada , f. , klypin , a. , & gottlber , s.2006 , mnras , 369 , 335 percival , w. j. , et al .
2007 , apj , 657 , 645 sanchez a.g . ,
baugh c.m . ,
percival w.j . , peacock j.a .
, padilla n.d . ,
cole s. , frenk c.s . , norberg p. , 2006 ,
mnras , 366 , 189 spergel , d. n. , et al .
2007 , apj , 170 , 377 stanford s.a .
, eisenhardt p.r . , dickinson m. , 1998 , apj , 492 , 461 szapudi i. , gaztaaga e. , 1998 , mnras , 300 , 493 tegmark , m. , et al .
2006 , phrvd , 74 , 123507 verde l. , et al . , 2002 , mnras , 335 , 432 vogeley , m. s. , geller , m. j. , park , c. , & huchra , j. p. 1994
, aj , 108 , 745 wechsler , r. h. , zentner , a. r. , bullock , j. s. , kravtsov , a. v. , & allgood , b. 2006 , apj , 652 , 71 white s.d.m . , 1979 , mnras 186 , 145 wild v. , et al . ,
2005 , 356 , mnras , 247 york d.g .
, et al . , 2000 , aj ,
120 , 1579 zehavi , i. , et al . , 2002 ,
apj , 571 , 172 zehavi , i. , et al . , 2004 ,
apj , 608 , 16 zehavi , i. , et al .
, 2005 , apj , 630 , 1 | we present measurements of the higher - order clustering of red and blue galaxies as a function of scale and luminosity made from the two - degree field galaxy redshift survey ( 2dfgrs ) .
we use a counts - in - cells analysis to estimate the volume averaged correlation functions , @xmath0 , as a function of scale up to order @xmath1 , and also the reduced void probability function .
hierarchical amplitudes are constructed using the estimates of the correlation functions : @xmath2 .
we find that : 1 ) red galaxies display stronger clustering than blue galaxies at all orders measured .
2 ) red galaxies show values of @xmath3 that are strongly dependent on luminosity , whereas blue galaxies show no segregation in @xmath3 within the errors ; this is remarkable given the segregation in the variance .
3 ) the linear relative bias shows the opposite trend to the hierarchical amplitudes , with little segregation for the red sequence and some segregation for the blue .
4 ) faint red galaxies deviate significantly from the `` universal '' negative binomial reduced void probabilities followed by all other galaxy populations .
our results show that the characteristic colour of a galaxy population reveals a unique signature in its spatial distribution .
such signatures will hopefully further elucidate the physics responsible for shaping the cosmological evolution of galaxies .
galaxies : statistics , cosmology : theory , large - scale structure . |
there is considerable interest in the study of equations of the form @xmath3 that describe breaking waves . in this paper
we consider the equation @xmath4 with parameters @xmath5 ( not all zero ) and @xmath6 .
this 4-parameter family contains several integrable equations . for @xmath7 and @xmath8 , equation reduces respectively to the camassa - holm equation @xcite @xmath9 and the degasperis - procesi equation @xcite @xmath10 while for @xmath11 , equation becomes the novikov equation @xcite @xmath12 the three equations , , are integrable in the sense of having a lax pair , a bi - hamiltonian structure , as well as hierarchies of local symmetries and local conservation laws , and they also possess peaked travelling wave solutions .
in addition to these integrable equations , many other non - integrable equations that admit breaking waves are included in the 4-parameter family .
for instance , there is the @xmath13-equation @xmath14 which unifies the camassa - holm and degasperis - procesi equations @xcite .
there is also a modified version of the @xmath13-equation @xcite @xmath15 which includes the novikov equation .
no other cases of the two equations and are known to be integrable @xcite . an equivalent form of the 4-parameter equation is given by @xmath16 in terms of the momentum variable @xmath17 with parameters @xmath18 this parametric equation is invariant under the group of scaling transformations @xmath19 , @xmath20 , @xmath21 with @xmath22 . in section [ conslaws ] , we classify the low - order conservation laws of equation and show that the hamiltonians of the camassa - holm and novikov equations are admitted as local conservation laws by equation if and only if @xmath23 and @xmath24 .
we consider peaked travelling waves in section [ peakons ] and use a weak formulation of equation to show that single peakon and multi - peakon solutions are admitted if and only if @xmath23 and @xmath25 when @xmath26 .
we derive the explicit equations of motion for @xmath27 peakon / anti - peakon solutions and also obtain the constants of motion inherited from the local conservation laws of equation . in section [ unifiedeqn ] ,
we combine the previous results to obtain a natural 1-parameter family of equations @xmath28 given by @xmath23 , @xmath24 , @xmath25 , @xmath26 , where a scaling transformation @xmath29 is used to put @xmath30 .
since this 1-parameter family unifies the camassa - holm and novikov equations , we will refer to it as the _ gchn equation_. ( similar unified equations have been considered previously from related perspectives @xcite . ) we then discuss some general features of the dynamics of its @xmath31 peakon / anti - peakon solutions and we show that wave - breaking occurs for certain solutions describing collisions between peakons and anti - peakons in the case @xmath2 .
finally , in section [ remarks ] , we make some concluding remarks including a possible scenario for wave - breaking in the cauchy problem for weak solutions .
for the 4-parameter equation , a _ local conservation law _
@xcite is a space - time divergence @xmath32 holding for all solutions @xmath33 of equation , where the _ conserved density _ @xmath34 and the _ spatial flux _
@xmath35 are functions of @xmath36 , @xmath37 , @xmath38 and derivatives of @xmath38 .
the spatial integral of the conserved density @xmath34 satisfies @xmath39 and so if the flux @xmath35 vanishes at spatial infinity , then @xmath40= \int_{-\infty}^{\infty } t dx=\const\ ] ] formally yields a conserved quantity for equation .
conversely , any such conserved quantity arises from a local conservation law .
if the conserved quantity is purely a boundary term , then the local conservation law is called _
trivial_. this occurs when ( and only when ) the conserved density is a total @xmath37-derivative and the flux is a total @xmath36-derivative , related by @xmath41 for all solutions @xmath33 of equation , where @xmath42 is some function of @xmath36 , @xmath37 , @xmath38 and derivatives of @xmath38 .
two local conservation laws are _ equivalent _ if they differ by a trivial conservation law , thereby giving the same conserved quantity up to boundary terms .
the set of all conservation laws ( up to equivalence ) admitted by equation forms a vector space on which there is a natural action @xcite by the group of all lie symmetries of the equation . for conserved densities and fluxes depending on at most @xmath43 , a conservation law can be expressed in an equivalent form by a divergence identity @xmath44 where @xmath45 is called the _ multiplier_. this identity is called the _
characteristic equation _
@xcite for the conserved density and flux . by balancing the highest order @xmath36-derivative terms @xmath46 on both sides of the equation
, we directly find that @xmath47 and @xmath48 . then balancing the terms @xmath49
, we see that @xmath50 .
hence the conserved density and the flux in the divergence identity must have the form @xmath51 its multiplier thus has the form @xmath52 in general , the differential order of a local conservation law is defined to be the smallest differential order among all equivalent conserved densities .
a local conservation law is said to be of _ low order _ if the differential orders of @xmath34 and @xmath35 are both strictly less than the differential order of the equation . consequently , conserved densities and fluxes of the form comprise all possible low - order conservation laws of equation .
the problem of finding all low - order conservations then reduces to the simpler problem of finding all low - order multipliers .
since equation is an evolution equation , it has no lagrangian formulation in terms of the variable @xmath38 .
in this situation , the problem of finding multipliers can be understood as a kind of adjoint @xcite of the problem of finding symmetries .
an _ infinitesimal symmetry _
@xcite of equation is a generator @xmath53 whose coefficient @xmath54 is given by a function of @xmath36 , @xmath37 , @xmath38 and derivatives of @xmath38 , such that the prolonged generator satisfies the invariance condition @xmath55 holding for all solutions @xmath33 of equation .
the lie symmetry group of equation is generated by infinitesimal symmetries with coefficients of the form @xmath56 if @xmath54 is at most linear in @xmath57 and @xmath58 , then the resulting generator will yield a group of point transformations @xcite , whereas if @xmath54 is nonlinear in @xmath57 or @xmath58 , then a group of contact transformations @xcite will be generated .
hence , all generators of lie symmetries admitted by equation are determined by the solutions of condition for @xmath59 .
( it is straightforward to solve this determining equation by maple to classify the lie symmetry group of equation , as shown in the appendix . )
the condition for determining all multipliers @xmath60 of low - order conservation laws admitted by equation consists of @xmath61 which arises from the property that the variational derivative ( euler operator ) @xmath62 annihilates an expression identically iff it is a space - time divergence @xcite
. this condition can be split with respect to @xmath63 and @xmath64-derivatives of @xmath63 , yielding an equivalent overdetermined system of equations on @xmath65 .
one equation in this system is given by the adjoint of the symmetry determining equation , @xmath66 holding for all solutions @xmath33 of equation .
solutions @xmath65 of this equation are called _ adjoint - symmetries _ ( or _ cosymmetries _ ) @xcite .
the remaining equations in the system comprise helmholtz conditions which are necessary and sufficient for @xmath65 to have the form . as a consequence ,
multipliers are simply adjoint - symmetries that have a certain variational form .
for any solution of the multiplier determining equation , a corresponding conserved density and flux of the form can be recovered either through integration @xcite of the characteristic equation , which splits with respect to @xmath67 , @xmath63 , @xmath68 , @xmath49 into a system of equations for @xmath34 and @xmath35 , or through a homotopy integral formula @xcite , which expresses @xmath34 and @xmath35 directly in terms of @xmath69 .
it is straightforward to show that @xmath34 and @xmath35 have the form of a trivial conservation law iff @xmath70 .
thus there is a one - to - one correspondence between equivalence classes of non - trivial low - order conservation laws and non - zero low - order multipliers .
both the camassa - holm equation and novikov equation possess low - order local conservations law given by the conserved densities @xcite @xmath71 where @xmath72 and @xmath73 , respectively , for the two equations .
in addition , the camassa - holm equation itself is a low - order local conservation law having the conserved density @xmath74 all of these conserved densities are related to hamiltonian structures for the two equations @xcite .
the corresponding multipliers are respectively given by @xmath75 to look for conserved densities of the same form for equation , we now classify all multipliers up to 1st - order @xmath76 as well as all 2nd - order multipliers with the specific form @xmath77 in each case it is straightforward to solve the determining equation by use of maple ( as shown in the appendix ) , which leads to the following classification result .
[ class - loworderq ] ( i ) equation admits 0th - order multipliers only in the following cases : @xmath78 ( ii ) for any @xmath6 and any @xmath79 , equation admits no 1st - order multipliers .
( iii ) equation admits 2nd - order multipliers of the form only in the following cases : @xmath80 in light of the adjoint relationship between multipliers and symmetries , the classification of 0th- and 1st- order multipliers in proposition [ class - loworderq ] is a counterpart of the classification of lie symmetries ( cf .
proposition [ class - pointsymm ] ) .
next we obtain the corresponding conserved densities and fluxes for each multiplier by first splitting the characteristic equation with respect to @xmath67 , @xmath63 , @xmath68 , @xmath49 where @xmath34 and @xmath35 have the form , and then integrating the resulting system of equations .
this yields the following low - order local conservation laws for equation .
[ class - lowordertx ] ( i ) the local conservation laws admitted by the wave - breaking equation with multipliers of at most 1st - order consist of three 0th - order conservation laws @xmath81 and two 1st - order conservation laws @xmath82 ( ii ) the local conservation laws admitted by the wave - breaking equation with 2nd - order multipliers of the form consist of two 2nd - order conservation laws @xmath83 in these conservation laws , any terms of the form @xmath84 in the case @xmath85 should be replaced by @xmath86 .
these conservation laws yield the following conserved integrals .
we start with the conservation laws at 0th order . from @xmath87 , we have @xmath88 which is the conserved mass for equation .
the conserved integral arising from @xmath89 is a weighted mass , @xmath90 interestingly , from @xmath91 we get a conserved integral which vanishes , but has a non - zero spatial flux .
this type of conservation law arises because the multiplier converts equation into the form of a total @xmath37-derivative .
next we look at the conservation laws at 1st order . from @xmath92 , the @xmath0 norm of @xmath33
is conserved , @xmath93 from @xmath94 , we have @xmath95 where @xmath96 is the center of mass of @xmath33 .
since @xmath97 is conserved , it can be evaluated at @xmath98 , which yields the relation @xmath99 .
this shows that the center of mass moves at a constant speed controlled by the @xmath0 norm of @xmath33 . finally , we consider the conservation laws at 2nd order . from @xmath100 ,
we get @xmath101 this shows that the @xmath102 norm of @xmath103 is conserved if @xmath104 does not change sign or if @xmath105 is an even integer .
the conserved integral arising from @xmath106 is a linear combination of the @xmath107 norms of @xmath108 as given by @xmath109 this can be written alternatively as a weighted @xmath110 norm when @xmath111 .
it is interesting to note that simultaneous conservation of both the @xmath0 and the weighted @xmath110 norms requires the condition @xmath112 which holds iff @xmath113 and @xmath114 , but in this case @xmath115 .
both the camassa - holm and novikov equations possess peaked travelling wave solutions @xcite , called peakons , @xmath116 where @xmath117 and @xmath72 , respectively , for the two equations .
peakons have attracted much attention in the study of breaking wave equations . in general , on @xmath118 ,
a peakon is a weak travelling wave solution satisfying an integral ( i.e. weak ) formulation of a breaking wave equation .
such a formulation is essential for deriving multi - peakon solutions .
however , single peakons can be derived directly from the travelling wave reduction of a breaking wave equation , which will be the approach we use here .
the manifest invariance of the 4-parameter equation under time - translation and space - translation symmetries implies the existence of travelling wave solutions @xmath119 where @xmath120 satisfies the ode @xmath121 for the travelling wave ode , an integral formulation is obtained through multiplying this ode by a test function @xmath122 ( which is smooth and has compact support ) and integrating over @xmath123 , leaving at most first derivatives of @xmath124 in the integral , which yields @xmath125 a weak solution of ode is a function @xmath120 that belongs to the sobolev space @xmath126 and that satisfies the integral equation for all smooth test functions @xmath127 with compact support on @xmath128 . to proceed we substitute a peaked travelling wave expression @xmath129 into equation and split up the integral into the intervals @xmath130 and @xmath131 .
the first term in equation yields , after integration by parts , @xmath132 similarly , the second term in equation gives @xmath133 provided @xmath134 so that the boundary terms at @xmath135 vanish .
the third and fourth terms in equation together yield @xmath136 when the terms are combined , we find that equation reduces to @xmath137 this equation is satisfied for all test functions @xmath122 iff @xmath138 which determines the amplitude @xmath139 in the peakon expression . thus we obtain the following result .
[ class - peakon ] the travelling wave equation admits a peakon solution only in the case @xmath140 where @xmath141 is the wave speed .
the resulting peakon solution of equation is given by @xmath142 when the nonlinearity power @xmath143 is a positive integer , then the wave speed is necessarily positive , @xmath144 , if @xmath143 is even , as in the case ( @xmath145 ) of the novikov equation , while if @xmath143 is odd , the wave speed can be either positive or negative , @xmath146 , as in the case ( @xmath147 ) of the camassa - holm equation .
the peakon solution satisfies equation only in the sense of a weak solution .
this means @xmath33 is a distribution in @xmath148 with respect to @xmath149 for some @xmath150 and in @xmath126 with respect to @xmath151 such that it satisfies the integral equation @xmath152 for all test functions @xmath153 in @xmath154 .
both the camassa - holm and novikov equations possess multi - peakon solutions @xcite which are a linear superposition of peaked travelling waves with time - dependent amplitudes and positions .
the form of these solutions is given by @xmath155 where the amplitudes @xmath156 and positions @xmath157 satisfy a hamiltonian system of odes @xmath158 given in terms of the hamiltonian function @xmath159 the poisson bracket @xmath160 in this system arises from the respective hamiltonian operator formulations @xcite of these two equations and has the standard canonical form in the case of camassa - holm equation and a certain non - canonical form in the case of the novikov equation .
we now investigate whether equation also admits multi - peakon solutions .
it will be convenient to use the notation @xmath161 where the summation is understood to go from @xmath162 to @xmath1 .
note that the @xmath37-derivatives of @xmath38 are given by @xmath163 and @xmath164 in terms of the sign function @xmath165 and the dirac delta distribution @xmath166 which has the properties @xmath167 for @xmath168 , and @xmath169 for all @xmath170 .
to begin , we substitute the general multi - peakon expression into the integral equation .
there are two ways we can then proceed .
one way is to assume @xmath171 at a fixed @xmath172 , split up the integral over @xmath37 into corresponding intervals , and integrate by parts , similarly to the derivation of the single peakon solution .
another way , which is simpler , is to employ the following result from distribution theory @xcite .
let @xmath173 be a piecewise @xmath174 function having at most jump discontinuities at a finite number of points @xmath175 in @xmath128 .
then , for any test function @xmath176 , @xmath177_{x_i } - \int_{-\infty}^{\infty } \psi \ns{f'}\;dx\ ] ] where @xmath178_{x_i } = f(x_i{}^+ ) - f(x_i{}^-)\ ] ] is the jump in @xmath173 at the point @xmath175 , and @xmath179 is the non - singular part of the distributional derivative of @xmath173 .
we will now use this integration by parts relation to evaluate each term in the integral equation . the first term in equation yields @xmath180_{\beta_i } - \psi(t,\beta_i)[u_{tx}]_{\beta_i } ) \ ; dt . \end{aligned}\ ] ] from expression , we see @xmath181 so thus @xmath182 holds a. e. in @xmath183 .
hence @xmath184 and thus we get @xmath185_{\beta_i } - \psi(t,\beta_i)[u_{tx}]_{\beta_i } ) \ ; dt .\ ] ] next , the second term in equation gives @xmath186_{\beta_i } \ ; dt .
\end{aligned}\ ] ] we now simplify the two parts of the integral involving @xmath187 . for the first part , we have @xmath188 after using relation and then integrating by parts . for the second part ,
since @xmath189 holds a. e. in @xmath183 , we have @xmath190_{\beta_i } \ ; dt - \iint_{-\infty}^{+\infty } \psi ( u^{p-1}\ns{u_x^2})_x \ ; dx\ ; dt \end{aligned}\ ] ] from applying the integration by parts relation . by simplifying @xmath191
a. e. with the use of relation , we see @xmath192 hence the integral becomes @xmath193_{\beta_i } \ ; dt .
\end{aligned}\ ] ] then we have @xmath194_{\beta_i } \ ; dt + \sum_{i}\int_{-\infty}^{+\infty } c\psi_x(t,\beta_i)u(\beta_i)^p [ u_x]_{\beta_i } \ ; dt .
\end{aligned}\ ] ] similarly , the third term in equation gives @xmath195_{\beta_i } \ ; dt . \end{aligned}\ ] ] we can simplify the integral involving @xmath122 by the same steps used for the previous integral .
this yields @xmath196 hence we then have @xmath197_{\beta_i } \ ; dt .
\end{aligned}\ ] ] finally , by combining the three terms , , with the fourth term in equation , we obtain @xmath198_{\beta_i } + c u(t,\beta_i)^p[u_x]_{\beta_i } ) \ ; dt \\&\qquad + \sum_{i } \int_{-\infty}^{+\infty } \psi(t,\beta_i ) ( -[u_{tx}]_{\beta_i } + \tfrac{1}{2}(pc - b)u(t,\beta_i)^{p-1}[u_x^2]_{\beta_i } ) \ ; dt .
\end{aligned}\ ] ] the jump terms are evaluated by @xmath199_{\beta_i } = 2\alpha_i\beta_i ' , \quad [ u_x]_{\beta_i } = -2\alpha_i \\ [ u_{tx}]_{\beta_i } = \frac{d\,[u_x]_{\beta_i}}{dt } = -2\alpha_i ' \\ [ u_x^2]_{\beta_i } = 2u_x(\beta_i)[u_x]_{\beta_i } = -4\alpha_iu_x(\beta_i)\end{gathered}\ ] ] which all follow directly from the expressions and .
thus we get @xmath200 this equation is satisfied for all test functions @xmath122 iff @xmath201 which determines the amplitudes @xmath202 and the positions @xmath203 in the multi - peakon expression .
thus we have established the following result .
[ class - npeakon ] the integral equation admits an @xmath1-peakon solution for all @xmath204 only in the case @xmath205 from propositions [ class - npeakon ] and [ class - peakon ] , we have a classification of all cases for which the 4-parameter equation possesses both single peakon and multi - peakon solutions .
[ class - peakons ] the 4-parameter equation admits single peakon and multi - peakon solutions iff @xmath206 in this case , a general @xmath1-peakon solution has the form , where the amplitudes @xmath156 and positions @xmath157 satisfy the system of odes @xmath207 in terms of the separations @xmath208 this result generalizes related work in ref.@xcite which established the existence of single and multi - peakon solutions for a 2-parameter equation defined by the case @xmath209 , @xmath30 of equation .
( in particular , the derivation in ref.@xcite was completely formal , whereas the steps here provide a rigorous proof applied to the more general 4-parameter equation . )
it is easy to check that the general @xmath1-peakon ode system reduces to the well - known multi - peakon systems for the @xmath13-equation when @xmath210 , which includes the the camassa - holm equation and the degasperis - procesi equation when @xmath211 and @xmath212 , respectively , as well as for the novikov equation when @xmath11 .
the ode system for the amplitudes and positions of the @xmath1 peakons in the expression inherits constants of motion ( i.e. time - independent quantities ) given by the conserved integrals that are admitted by equation in the case . from theorem [ class - lowordertx ] , there are six conserved integrals which we can consider .
the first conserved integral yields @xmath213 this quantity @xmath214 is the total mass for the @xmath1-peakon solution .
a weighted mass arises from the second conserved integral , @xmath215 the next conserved integral gives @xmath216 which is the @xmath0 norm of the @xmath1-peakon solution .
the fourth conserved integral does not exist in the case , since @xmath217 and @xmath218 together imply that @xmath219 .
last , the two conserved integrals and are nonlinear in @xmath220 which is a distribution . as a consequence ,
both these integrals are ill - defined for the the @xmath1-peakon solution .
from theorems [ class - lowordertx ] and [ class - peakons ] , the low - order conservation laws as well as the @xmath1-peakon solution expression of the camassa - holm and novikov equations are admitted simultaneously by the 4-parameter equation iff its parameters @xmath221 satisfy @xmath222 after a scaling transformation @xmath29 is used to put @xmath30 , equation reduces to the 1-parameter gchn equation presented in section [ intro ] .
the explicit system describing @xmath1-peakon solutions of the gchn equation for all @xmath26 is given by @xmath223 where @xmath156 and @xmath157 are , respectively , the amplitudes and positions appearing in the general @xmath1-peakon expression . the @xmath0 norm of the @xmath1-peakon solution provides a constant of motion @xmath224 which is determined by the initial amplitudes and initial separations .
when all of the amplitudes are positive , @xmath225 , for all @xmath226 , the solution expression is a superposition of @xmath27 peakons , each of which is right moving . in this case , the constant of motion directly gives the inequality @xmath227 , @xmath228 , which implies that any collisions among the @xmath1 peakons are elastic .
when all of the amplitudes are negative , @xmath229 , for all @xmath226 , the solution expression is instead a superposition of @xmath27 anti - peakons , each of which is either right moving if @xmath143 is even or left moving if @xmath143 is odd .
similarly to the previous case , the constant of motion yields @xmath230 , @xmath228 , implying that any collisions among the @xmath1 anti - peakons are elastic . in the case when some amplitudes have opposite signs , or an amplitude changes its sign at some @xmath172 ,
the solution expression then describes a superposition of both peakons and anti - peakons .
although the constant of motion is still non - negative , the amplitudes are no longer bounded by @xmath231 . as a consequence
, wave breaking can occur in collisions , which we will now show for the case @xmath2 . for @xmath2 ,
the system describing @xmath232-peakon solutions @xmath233 takes a simple form .
first , the constant of motion can be used to express the relative separation @xmath234 in terms of the two amplitudes @xmath235 and @xmath236 through the relation @xmath237 then , the equations of motion for the two positions @xmath238 and @xmath239 and the two amplitudes @xmath235 and @xmath236 are given by @xmath240 with @xmath241 and @xmath242 if another constant of motion could be found for this system , then the system could be reduced to two separated odes for the two amplitudes , plus two quadratures for the two positions , which would allow the general solution to be obtained . even without another constant of motion ,
it is still possible to do a qualitative analysis of all solutions by studying the phase plane @xmath243 of the coupled odes for the amplitudes .
we start from the relation , which imposes inequalities on the amplitudes , @xmath244 for a given value of @xmath245 , these two inequalities define the domain for all @xmath232-peakon solutions in the phase plane @xmath243 .
the boundary of the domain corresponds to the two equalities latexmath:[\[\label{h - circ } \alpha_1 ^ 2 + \alpha_2 ^ 2 = h , \quad ( \alpha_1 + \alpha_2)^2 = h , \quad the circle comprises the equilibrium points of the amplitude odes in the phase plane .
each point on the circle is a limit of a @xmath232-peakon solution describing an asymptotic superposition of two @xmath162-peakon solutions , in which the amplitudes are constant and the positions are infinitely separated .
the lines each constitute a degenerate @xmath232-peakon solution in which the two positions coincide and the sum of the two amplitudes is constant , describing a peakon solution @xmath247 in the case of the upper line , and an anti - peakon solution @xmath248 in the case of the lower line .
the entire solution domain divides into four parts which are related by a reflection symmetry @xmath249 .
one part of the domain is given by the points lying between the circle and the upper line in the first quadrant , which comprises all solutions describing two peakons .
there is a counterpart given by the points lying between the circle and the lower line in the third quadrant , which comprises all solutions describing two anti - peakons .
the two other parts of the domain comprise all solutions describing a peakon and an anti - peakon .
these parts are given by the points between the segments of the upper and lower lines that lie outside of the circle .
within this solution domain in the phase plane , the flow defined by the amplitude odes depends on the nonlinearity power @xmath143 and the sign of the separation @xmath250 .
we are interested in flows that describe a collision between a peakon and an anti - peakon .
this condition can be used to determine @xmath251 at @xmath98 at each point in the phase plane by considering the ode @xmath252 for the separation . if @xmath253 , then the relative separation @xmath254 between the peakon and anti - peakon will be decreasing only if @xmath255 . similarly , if @xmath256 , then the relative separation @xmath254 between the peakon and anti - peakon will be decreasing only if @xmath257 .
hence , a necessary condition for a collision to occur is that @xmath258 and @xmath250 have opposite signs during the flow .
since @xmath259 can occur only on the upper and lower lines , which are boundaries of the domain in which solutions describe a collision between a peakon and an anti - peakon , we can impose @xmath260 at each point in the phase plane .
note @xmath261 holds iff @xmath262 when @xmath143 is odd , and @xmath263 when @xmath143 is even .
the points given by @xmath262 in the phase plane consist of the lines and @xmath264 , while the points given by @xmath265 consist of the lines that are perpendicular to each of those three lines .
consequently , hereafter we will consider initial conditions @xmath266 and @xmath267 without loss of generality .
( note that reversing the sign in the initial condition will correspond to reflecting the flow about the line @xmath268 in the phase plane . ) under the collision condition and initial conditions
, the flow then depends only on the nonlinearity power @xmath143 .
the case @xmath147 , which represents the camassa - holm equation , is special , since there is another constant of motion @xmath269 which is given by the total mass .
this implies that the flow simply consists of parallel lines in the phase plane . in all other cases
@xmath270 , the flow is no longer given by straight lines and has a much richer structure . ] ] the flows for all even powers @xmath271 are qualitatively similar to the case @xmath145 , which represents the novikov equation . a picture of the phase plane for @xmath272 is shown in fig .
[ p=4-plane ] . clearly , in the second quadrant , the upper line @xmath273 is a stable asymptotic attractor for solutions describing a peakon ( @xmath274 ) and an anti - peakon ( @xmath275 ) , while the lower line @xmath276 is an unstable asymptotic attractor . in the fourth quadrant
, these behaviours are reversed .
the flows for all other odd powers @xmath277 are qualitatively similar to the case @xmath278 which is shown in fig .
[ p=3-plane ] . in the second quadrant ,
both the upper and lower lines @xmath279 are stable asymptotic attractors for solutions describing a peakon ( @xmath274 ) and an anti - peakon ( @xmath275 ) .
the line @xmath280 is an unstable asymptotic attractor . in the fourth quadrant ,
the behaviour is the same . in all cases
@xmath270 , the flow will evolve the initial amplitudes toward a stable attractor line .
this evolution is shown in figs .
[ p=3-collide ] and [ p=4-collide ] for the cases @xmath278 and @xmath272 , respectively , where the initial positions of the peakon and anti - peakon are chosen to be distinctly separated .
we see that the peakon and anti - peakon collide such that their peak amplitudes become closer while the slope at locations @xmath37 in between the two peaks rapidly increases ( without bound ) as relative separation between their positions decreases to zero in a finite time .
this blow - up in the slope seen in figs .
[ p=3-blowup ] and [ p=4-blowup ] is an example of wave breaking .
.35 ( @xmath281 , @xmath282),title="fig : " ] .35 ( @xmath281 , @xmath282),title="fig : " ] .35 ( @xmath281 , @xmath282),title="fig : " ] .35 ( @xmath281 , @xmath282),title="fig : " ] .33 ( @xmath283 , @xmath284),title="fig : " ] .33 ( @xmath283 , @xmath284),title="fig : " ] .33 ( @xmath283 , @xmath284),title="fig : " ] .33 ( @xmath283 , @xmath284),title="fig : " ] there is a qualitative explanation of why the blow - up in the slope @xmath58 between the two peaks in a collision solution occurs in a finite time .
consider the asymptotic attractor solution @xmath285 corresponding to the upper line .
this solution arises from the initial condition @xmath286 and @xmath287 .
the amplitude odes
yield @xmath288 whereby @xmath289 and @xmath290 as @xmath291 such that @xmath292 is constant for all @xmath226 .
any solution having an initial condition close to @xmath286 and @xmath287 will exhibit a similar long - time behaviour for @xmath235 and @xmath236 , as a consequence of continuous dependence of solutions on initial data for the odes . since @xmath293 , the solution remains continuous and bounded at all @xmath37 for @xmath226 , whereas the slope @xmath294 has jump discontinuities at @xmath295 and @xmath296 and becomes unbounded at @xmath297 ( with @xmath298 ) as @xmath299 .
the same kind of wave - breaking behaviour can be expected to occur in collisions between peakons and anti - peakons when @xmath300 . .3 , title="fig : " ] .3 , title="fig : " ] .3 , title="fig : " ] .3 , title="fig : " ] .3 , title="fig : " ] .3 , title="fig : " ] .3 , title="fig : " ] .3 , title="fig : " ]
at first sight , the gchn equation seems closely analogous to the @xmath13-equation : both equations unify two integrable equations , possess @xmath1-peakon solutions , and exhibit wave breaking phenomena . however , there are important differences .
firstly , the nonlinearities in the @xmath13-equation are purely quadratic , whereas the gchn equation has nonlinearities of degree @xmath301 and thereby connects two integrable equations with different nonlinearities .
secondly , the @xmath0 norm of solutions @xmath33 is conserved for the @xmath13-equation only if @xmath302 , when the @xmath13-equation reduces to the camassa - holm equation .
in contrast , the @xmath0 norm is conserved for the gchn equation for all @xmath6 . in a subsequent work
, we will explore further properties of the gchn equation and its multi - peakon solutions .
there are numerous interesting questions .
can a wave - breaking result similar to those for the camassa - holm and novikov equations be established for classical solutions ? how will the wave - breaking behaviour depend on @xmath143 ?
in particular , a plausible criteria for wave - breaking is @xmath303 which generalizes the criteria known @xcite in the camassa - holm case @xmath147 and the novikov case @xmath145 . in another direction , for any @xmath143 other than these two known integrable cases @xmath147 and @xmath145 , does the equation have a hamiltonian formulation or perhaps integrability properties ?
s.c . anco is supported by an nserc research grant . p.l .
da silva and i.l .
freire would like to thank fapesp ( scholarship n. 2012/22725 - 4 and grant n. 2014/05024 - 8 ) and capes for financial support .
freire is also partially supported by cnpq ( grant n. 308941/2013 - 6 ) .
the referee is thanked for remarks which have improved this paper . to classify all of the lie symmetries admitted by the 4-parameter equation
, we first substitute a general coefficient function @xmath59 into the symmetry determining equation .
next we eliminate @xmath63 , @xmath304 , @xmath305 through writing the equation in the solved form @xmath306 and doing the same for its differential consequences .
the determining equation then splits with respect to @xmath220 , @xmath307 , @xmath49 , @xmath68 , @xmath67 , @xmath308 into a linear overdetermined system of 10 equations for @xmath309 : @xmath310 equation shows that @xmath54 is a linear function of @xmath57 and @xmath58 , and hence @xmath311 for some functions @xmath312 . after simplifying the remaining equations , we obtain a system of 14 equations @xmath313 we solve this linear overdetermined system by the following steps .
first , an integrability analysis of the system of equations is carried out using the maple package _ rifsimp _ , which yields 9 cases .
next , in each case the reduced system of equations is integrated using the maple command _
pdsolve_. last , the solutions are merged , which leads to the following classification result .
[ class - pointsymm ] ( i ) for any @xmath6 and any @xmath79 , equation admits no contact symmetries .
( ii ) the point symmetries admitted by equation for all @xmath6 and all @xmath79 consist of @xmath314 ( iii ) equation admits additional point symmetries only in the following cases : @xmath315 the symmetries @xmath316 , @xmath317 , @xmath318 respectively generate one - dimensional point transformation groups consisting of time - translations @xmath319 , space - translations @xmath320 , and scalings @xmath321 , @xmath322 , with group parameter @xmath323 . the extra symmetry @xmath324 generates the one - dimensional point transformation group @xmath325 , @xmath326 , which is a galilean boost , and the other extra symmetry @xmath327 generates the one - dimensional point transformation group @xmath328 , @xmath329 , which is a non - rigid dilation .
to classify all 1st - order multipliers admitted by the 4-parameter equation , we first substitute the expression into the determining equation , which splits into a linear overdetermined system of 5 equations for @xmath330 .
the system contains the equations @xmath331 which yield @xmath332 after the remaining 3 equations are split with respect to @xmath57 and @xmath58 , we obtain the following system of 8 equations @xmath333 we solve this linear overdetermined system by the same three steps used in solving the symmetry system . this yields the five distinct cases presented in parts ( i ) and ( ii ) of proposition [ class - loworderq ] . finally , by splitting and simplifying the determining equation for second - order multipliers of the form , we obtain a linear overdetermined system of 13 equations for @xmath334 .
one of the equations in this system is given by @xmath335 which yields @xmath336 the remaining 10 equations then split with respect to @xmath58 , leading to a system of 6 equations @xmath337 solving this linear overdetermined system by the same steps used in solving the multiplier system , we obtain the two distinct cases presented in part ( iii ) of proposition [ class - loworderq ] . p.l .
da silva and i.l .
freire , an equation unifying both camassa - holm and novikov equations , _ proceedings of 10th aims conference _ , ( 2014 ) [ accepted ] .
see also p.l .
da silva and i.l .
freire , strict self - adjointness and shallow water models , _ arxiv:1312.3992 _ ( 2013 ) . | a 4-parameter polynomial family of equations generalizing the camassa - holm and novikov equations that describe breaking waves is introduced .
a classification of low - order conservation laws , peaked travelling wave solutions , and lie symmetries is presented for this family .
these classifications pick out a 1-parameter equation that has several interesting features : it reduces to the camassa - holm and novikov equations when the polynomial has degree two and three ; it has a conserved @xmath0 norm and it possesses @xmath1-peakon solutions , when the polynomial has any degree ; and it exhibits wave - breaking for certain solutions describing collisions between peakons and anti - peakons in the case @xmath2 . |
decay of superdeformed ( sd ) rotational bands out to normal deformed ( nd ) states is one of the most interesting nuclear structure problems .
it can be viewed as a shape - coexistence phenomena , and the mixing mechanism of two states having very different internal structures can be studied as functions of angular momenta and excitation energies . using the compound mixing model @xcite ,
we have investigated the decay - out phenomena in @xcite , where a consistent description was presented and the rapid decay - out was well understood .
after the calculation of @xcite , more systematic and improved calculations have been performed @xcite , in which all the relevant quantities to the decay - out have been calculated without any adjustable parameters .
we have found , however , it is rather difficult to reproduce the decay - out spin for individual sd bands ( cf . figs .
[ fig : a150 ] , [ fig : a190 ] ) . the basic idea of our framework @xcite is based on a simple two - well - mixing : the two groups of unperturbed states belonging to wells , i ( e.g. nd ) and ii ( e.g. sd ) , are separated by a potential barrier in some collective coordinate space ( e.g. deformations ) .
the most important is the ( mean ) coupling strength between the two , and is estimated as @xcite @xmath1 where @xmath2 is the frequency of each well , @xmath3 is the mean level distance , and @xmath4 is the transmission coefficient of the tunneling process between the wells with the least action @xmath5 .
depending on the situation of states in each well , one of three cases should be chosen in eq .
( [ eq : vcoupl ] ) : an example of ( 1a ) is the decay of high-@xmath6 isomers @xcite , in which severe breakdown of @xmath6-hindrance has been observed recently .
the present problem of decay - out of near - yrast sd bands corresponds to ( 1b ) , where the nd states are compound states and described by the goe model @xcite . in higher energies thermally
excited sd bands are rotationally damped @xcite , and ( 1c ) may be most appropriate for the decay - out of such sd continuum states @xcite . in the present case ,
the relative intensity of @xmath7-rays inside the sd band , which is calculated by the total decay - out probability @xmath8 at spin @xmath9 , is almost the only observable and will be discussed in the followings . according to @xcite
, @xmath10 is determined by combinations , @xmath11 and @xmath12 , of four quantities ; the spreading width @xmath13=@xmath14 due to the coupling ( [ eq : vcoupl ] ) , the level density @xmath15 of the nd compound states , and the @xmath7-decay widths @xmath16 and @xmath17 of the sd and nd bands , respectively , where @xmath16 is of rotational e2 , while @xmath17 is mainly of statistical e1 ( see @xcite for details ) .
it is worth mentioning that the model of @xcite for @xmath10 was re - examined by means of a statistical model of compound nucleus @xcite , and both models were found to give identical results for actual range of four parameters being relevant to decay of both the @xmath18 and 190 sd nuclei . in this talk
, we present the result of systematic calculations for the relative @xmath7-ray intensity as a function of angular momentum in comparison with experimental data .
we also discuss how the decay - out spin is understood from the calculations .
possible directions for improvements of the theoretical framework are suggested further .
quantities to be calculated are @xmath13 , @xmath19 , @xmath16 , and @xmath17 as functions of spin .
@xmath16 is the usual rotational e2 width , while the level density and the statistical e1 width are given by @xmath20 in unit of mev , where @xmath21 is excitation energy of the sd band from the nd yrast state , and @xmath22 parameter is taken from an empirical analysis of @xcite including the shell and temperature effects .
the decay - out spin @xmath23 in the @xmath24 region is generally low so that the back - shift is used for @xmath24 nuclei , i.e. @xmath21 is replaced by @xmath25 mev in @xmath26 , throughout the present investigation . as for @xmath13 , the least action path in the @xmath27-plane
is solved and the frequency @xmath28 and the action @xmath29 along it are calculated
. therefore all what we need are excitation energies of both sd and nd rotational bands , their quadrupole moments , and the potential energy surface and the mass tensor .
we have used the nilsson - strutinsky calculation for potential with the pairing correlations included in the rpa order , and the pairing hopping model for mass tensor @xcite ; see @xcite for the detailed formulation . in figs .
[ fig : a150 ] and [ fig : a190 ] the results of calculation for relative intensities are shown , where those for the lowest band in each parity and signature ( four bands ) are included in one nucleus .
the basic characteristics of intensity pattern are reproduced in both the @xmath18 and 190 regions ; especially the rapid decrease of transitions at lower spins .
apparently , however , the decay - out spin does not agree precisely , and the detailed features , like the relative ordering of decay - out for excited bands in one nucleus , or relative difference of neighbouring nuclei , are not well described . comparing figs .
[ fig : a150 ] and [ fig : a190 ] , decay - out spins are rather well reproduced on average for the @xmath18 nuclei , while the calculated @xmath23 s are still higher than the observed ones for the @xmath24 nuclei , even though the back - shift is used for them . although there are some cases where the yrast sd band decays at higher spin than the excited ones in figs .
[ fig : a150 ] and [ fig : a190 ] , it is because the energy ordering is determined in the feeding - spin region ; the ` real ' yrast band survives longest in most cases . as is shown in figs .
[ fig : a150 ] and [ fig : a190 ] , the decay - out spin is not precisely reproduced in individual cases : what are the reasons ?
possible sources of ambiguity in our model may be the level density and the mass tensor ; we believe that the potential energy surface by the nilsson - strutinsky method is reliable . as for the change of the level density
@xcite it was found that , for example , a factor about @xmath30 is enough to fit @xmath23 in @xmath31dy , while factor @xmath32 is required for @xmath33hg ; the latter may be out of the range of allowed ambiguity .
the mass tensor is calculated by using the fermi - gas estimate of the number of level - crossings @xcite , and detailed properties of individual nucleus is not included .
therefore , we try to use a mass tensor multiplied by a scaling factor @xmath34 and adjust it so as to fit @xmath23 .
resultant intensities for four selected nuclei are shown in fig .
[ fig : fitmass ] , and the values of @xmath34 and of various quantities at @xmath23 are summarized in table [ tab : gamma ] .
again , the correction factor @xmath34 for @xmath18 nucleus is rather small but is considerably larger for @xmath24 nucleus . .
calculated quantities at the decay - out spin which is fitted by scaling the mass ( fig .
[ fig : fitmass ] ) . [ cols="^,^,^,^,^,^,^,^,^,^,^",options="header " , ] in order to see how the decay - out spin is determined by the four relevant quantities , we note that @xmath35 holds in realistic situations @xcite .
two quantities , @xmath36 and @xmath12 , are plotted as functions of spin in fig .
[ fig : gamdsni ] .
@xmath10 takes appreciable value when @xmath37 , i.e. two curves crosses in fig .
[ fig : gamdsni ] , and @xmath23 is roughly determined by their crossing point .
moreover , the decay - out occurs more rapidly as the crossing angle between the two curves gets larger .
@xmath38 decreases more rapidly in the low - spin region because of the @xmath39 behaviour of the rotational e2 transitions of sd bands , and therefore the decay - out tends to be more rapid when @xmath23 is lower ; this behaviour is well observed in figs .
[ fig : a150 ] and [ fig : a190 ] .
@xmath12 takes similar values for sd bands within the @xmath18 or 190 region , while @xmath36 takes variety of values mainly due to the tunneling probability , which reflects the different potential landscapes depending on configurations and nuclei .
this is the main reason why the range of calculated @xmath23 spreads rather broadly in contradiction to experimentally observed trend .
in contrast , observed hindrance factors for the decay of @xmath6-isomers take broad range of values in accordance with the results of similar calculations of tunneling @xcite .
the observed trend that @xmath23 falls in a rather narrow range in both the @xmath18 and 190 regions is not easy to understand .
finally , we have recently developed a microscopic model of thermally excited sd states that describes both the barrier penetration leading to the decay - out to the nd states , and the rotational damping causing fragmentation of rotational e2 transitions @xcite .
as an example , @xmath40 , the number of rotational sd bands in the quasi - continuum @xmath7-ray region for @xmath41eu is shown in fig .
[ fig : nband ] , where solid ( dashed ) curve is a result using eq .
( 1c ) ( eq .
inclusion of the barrier penetration reduces @xmath40 dramatically in the low - spin region and the calculation nicely agrees with experimental data @xcite , which clearly shows that our model is capable for the unified understanding of feeding and decay of the sd bands .
systematic comparison of the calculated and observed relative intensities have been performed , and it is found that basic features of the decay - out of sd bands are well understood within our theoretical framework .
however , the decay - out spin is not precisely reproduced in the calculation in individual cases .
especially an average value of @xmath23 for the @xmath24 nuclei is somewhat overestimated , and the range of calculated @xmath23 s seems too wide in both the @xmath18 and 190 regions .
it should , however , be noticed that the average value of calculated @xmath23 in the @xmath24 region is much smaller than that in the @xmath18 region by about 10@xmath42 ; the difference in the experiment is even more , about 15@xmath42 , though . as a possible source of improvement
a simple scaling of the mass tensor is considered .
we believe that microscopic calculations of mass tensor based on counting the number of level crossings has to be done in order to include detailed nuclear structure effects in each sd band . for the detailed comparison ,
the reproduction of excitation energies of the sd bands is very important , which is not enough in the present calculations ; for example , that of @xmath43pb in table [ tab : gamma ] is about 1 mev higher than the experimentally deduced value . for this purpose ,
the improvement of the potential energy surface may also be necessary , e.g. by making use of an another mean - field potential like woods - saxon or the one obtained by skyrme hartree - fock method .
these are remaining future problems . as an another possible source of improvement
, modification of the compound mixing model @xcite may be considered . at higher spins where the excitation energy of sd bands is not so high , the nd states to which the sd band couples are not chaotic , and then strong reduction of mixing amplitudes
may be expected @xcite .
this reduction leads to a delay of decay - out , which hopefully helps to decrease @xmath23 especially in the @xmath24 .
there are , however , some ambiguities of the amount of reduction , and a precise modeling of this effect is necessary to draw a definite conclusion .
it is also an interesting future problem . | decay of the superdeformed bands have been studied mainly concentrating upon the decay - out spin , which is sensitive to the tunneling probability between the super- and normal - deformed wells .
although the basic features are well understood by the calculations , it is difficult to precisely reproduce the decay - out spins in some cases .
comparison of the systematic calculations with experimental data reveals that values of the calculated decay - out spins scatter more broadly around the average value in both the @xmath0 150 and 190 regions , which reflects the variety of calculated tunneling probability in each band . |
as well as applications to statistical physics , conformal field theory has proved to be a very powerful tool in string theory . in particular , the ground state of a critical string corresponds to a conformal field theory with a specific central charge .
it is of particular interest to classify all such ground states which can therefore be done by finding the space of all conformal field theories of a given central charge .
this `` moduli space '' forms the space of string vacua and may be considered as the stringy analogue of the space of einstein metrics in general relativity .
the moduli space of conformal field theories thus gives rise to two immediate applications .
firstly one may try to gain an understanding of stringy effects in quantum gravity by comparing the moduli space of conformal field theories with the space of einstein metrics for a given class of backgrounds .
secondly one may assume that space - time is in the form of flat four - dimensional minkowski space times some compact part @xmath7 .
the space of possible @xmath7 s leads to a space of theories of particle physics ( i.e. , particle masses , couplings , etc . ) in four dimensional space time ( see , for example , @xcite ) . in this latter case
@xmath7 has a euclidean signature . because of the difficulty in analyzing conformal field theories associated to a target space with indefinite signature we will need to restrict our attention to the latter scenario .
it should be expected however that many of the features we observe in these lectures should carry over to the former case of stringy quantum gravity of all of space - time . in section [ s : cft ] we will deal with simple examples of non - supersymmetric conformal field theories and their moduli space to introduce the basic concepts we will require later in these lectures .
the basic example central to a great deal of work in this subject will be that of @xmath8 theories and the linear sigma model whose target space is a circle .
the notion of duality appears here and will be of some interest later in these lectures .
we will find that extending our ideas to more complicated examples is very difficult to achieve in general . because of this we are forced to impose restrictions on the type of conformal field theories we study .
in particular we want to focus on conformal field theories which are associated to some geometric target space ( or perhaps some slightly generalized notion thereof ) .
we also impose that the conformal field theory has @xmath9=2 supersymmetry .
the effect of this is to force the target space to be a space with a complex structure . in terms of the flat four - dimensional minkowski space point of view
these conditions amount the existence of a space - time supersymmetry . for the purposes of these lectures we may simply regard these conditions as providing us with enough structure to use the tools of algebraic geometry . in section [ s : torus ] we will study the superconformal field theory for a sigma model with a complex one - torus as the target space .
this will allow us to introduce the complex coordinates which prove to be extremely useful for dealing with later examples .
section [ s : k3 t ] will cover briefly the case of a k3 surface as the target space .
in this case we have @xmath9=4 supersymmetry .
this section will also introduce the concept of a `` blow - up '' which is a key construction in algebraic geometry and thus also appears naturally in the context of superconformal field theories .
this blow - up also appears to be of central importance to understanding some global issues of the moduli space of @xmath9=2 theories and so it will become something of a recurring theme in later sections . in the sections discussed thus far we will find that using a metric as an effective description of the target space suffices . for the rest of the lectures however we will study examples which require more radical approaches . in particular
we will be required to think in terms of algebraic geometry rather than differential geometry .
for the cases we discuss in the later sections , the moduli spaces factor into two parts @xmath10 ( moduli some discrete symmetries and so long as we are careful about the boundary points ) . in geometric terms @xmath11 corresponds to deformations of the ( complexified ) khler form on @xmath7 and @xmath12 corresponds to deformations of the complex structure of @xmath7 .
the factor @xmath12 turns out to be simple to understand and may be analyzed classically . in order to understand the structure of the moduli space of a particular class of conformal field theories we will have to give three interpretations to each point in @xmath11 : 1 . the desired interpretation as a theory with some target space @xmath7 with a specific khler form .
this is the most difficult to analyze .
2 . a theory with some flat target space containing @xmath7 with a specific khler form . in some limit
the fields in this theory are required to live in @xmath7 .
this is the `` linear '' of @xcite .
3 . a theory with some space @xmath13 , related to @xmath7 by `` mirror symmetry '' , where the point in moduli space specifies a complex structure on @xmath13 . we will find that the third interpretation in terms of @xmath13 provides the simplest context in which to compute the moduli space but that we require the linear as an intermediary to translate between interpretations on @xmath7 and @xmath13 for each point in this space .
in section [ s : d=3 ] we will look at the simplest non - trivial example of the above and explicitly compute @xmath11 . in section [ s : phase ]
we will consider the more general case .
finally in section [ s : conc ] we present a few concluding remarks .
we will begin our discussion with the simplest . for further details and references as well as
an excellent introduction to conformal field theory the reader is referred to @xcite . consider a field theory in a two - dimensional space @xmath14 whose action is given by @xmath15 where @xmath16 is real and we are using complex coordinate @xmath17 and its conjugate @xmath18 on the world - sheet , @xmath14 .
classically this action is conformally invariant and under quantization one indeed obtains a conformal field theory with central charge @xmath8 .
that is , the stress tensor @xmath19 which in this case is proportional to the normal ordered product @xmath20 obeys the operator product expansion @xmath21 with @xmath8 . in this way
we say that we have a @xmath8 conformal field theory description of the real line @xmath22 .
a simple generalization of ( [ eq : bsm ] ) is to give the field @xmath16 an index @xmath23 which runs @xmath24 .
that is , the value of @xmath25 lies in @xmath26 .
we may also think of @xmath16 as defining a map @xmath27 .
the stress tensor is now in the form @xmath28 and we obtain ( [ eq : vir ] ) with @xmath29 .
thus , at least in this simple case we see that the central charge is a measure of the dimension of the target space of the .
returning to the case @xmath30 , we may make an alternative simple generalization of our simple model by imposing periodic boundary conditions on @xmath16 .
that is , we consider the target space to be a circle of radius @xmath31 which implies @xmath32 .
actually it is convenient to rescale @xmath16 such that @xmath33 .
then the action is @xmath34 this is a conformal field theory with @xmath8 for any value of @xmath31 agreeing with the idea that a circle is a one - dimensional object irrespective of its size !
we have a family of field theories parametrized by @xmath31 .
the key purpose of these lectures is to discuss such families of conformal field theories .
the moduli space of theories is the space mapped out by the parameters in the theory .
thus the moduli space for theories on a circle appears at first sight to be the real half line @xmath35 mapped out by @xmath36 .
certainly this is the moduli space for circles . in order to be sure that we have the right moduli space
however it is important to ask the following question .
do two distinct points in the supposed moduli space actually correspond to identical conformal field theories ? to answer this question we need to know more about the theory than just the action ( [ eq : bsmr ] ) .
while different values of @xmath31 certainly give different actions it may be the case that when we work out the spectrum of fields and their correlators we end with identical field theories with different values of @xmath31 . thus turns out frequently to be the case as we shall see .
consider the local structure of the moduli space around a point corresponding to the action @xmath37 .
for a field @xmath38 in this theory we may build a new field theory @xmath39 for some small @xmath40 . to maintain conformal invariance
we require that the term added be conformally invariant .
this implies that @xmath41 must be of weight ( dimension ) @xmath42 with respect to @xmath43 and @xmath18 .
such an operator is called `` marginal '' .
it is important to realize that the operator @xmath41 belongs to the field theory given by @xmath37 and not @xmath44 . in order that we may `` transport '' this field to a new one in the new theory which is also marginal we require that @xmath41 should not interact with itself to cause corrections to its own weight @xcite .
such a field is called `` truly marginal '' . such fields naturally span the tangent space to any point in the moduli space .
it turns out that something rather special happens with @xmath45 ( see @xcite for more details ) .
the resultant conformal field theory is completely equivalent to the field theory that corresponds to a string propagating on the group manifold @xmath46 .
this has two striking consequence .
firstly @xmath46 is three - dimensional and @xmath8 thus ruining our initial hope that the central charge might give the dimension of the target space .
secondly , the extra symmetries given by the affine algebra of @xmath46 map @xmath41 to @xmath47 for this theory .
thus decreasing @xmath31 appears to be the same as increasing @xmath31 away from this point in the moduli space .
actually what we see here is the famous duality symmetry for a string on a circle @xcite .
it turns out that the partition function of the string on a circle of radius @xmath31 is identical to that of a string on a circle of radius @xmath48 .
the reason for this is quite simple to picture .
as in normal quantum mechanics , the momentum for the string going around the circle is quantized just like a `` particle in a box '' .
this leads to `` momentum mode '' eigenstates with energy proportional to @xmath49 for @xmath50 . unlike the quantum theory of a particle we also have `` winding modes '' where the string wraps around the circle .
clearly the string must wind an integer number , @xmath51 , times round the circle forcing these modes to be quantized too .
the energy of these modes goes like @xmath52 .
thus , the symmetry appears as an exchange in the rles of @xmath53 and @xmath51 , i.e. , an exchange of winding modes with momentum modes .
( note that in order to prove the equivalence of two conformal field theories it is not sufficient just to show that partition functions are identical . in this case
however we know that this duality holds near the point @xmath45 because of the symmetry @xmath54 and we may integrate along this marginal direction to extend the symmetry to a whole @xmath55-symmetry acting on the line of @xmath31 s
. this @xmath55 symmetry must be the duality since this is the only symmetry of the partition function with a fixed point at @xmath45 . ) at this point our moduli space appears to be the half real line closed at both ends by @xmath56 with the point at infinity corresponding to @xmath22 , i.e. , a circle of infinite radius .
actually this is not the full moduli space of @xmath8 conformal field theories .
we may form orbifolds of theories we already have . to form an orbifold we divide the theory out by a discrete symmetry
this not only projects out the non @xmath57-invariant states from the original theory but also adds in `` twisted '' modes corresponding to open strings in the original theory whose ends are identified under @xmath57 .
the circle generically admits a @xmath55 symmetry by @xmath58 .
this leads to a new line of theories .
we will have more to say about orbifolds later .
one may also divide the @xmath46 group manifold at the @xmath45 point by any discrete subgroup of @xmath46 .
most of these subgroup lead to theories already accounted for but the three exceptional subgroups ( the bitetrahedral , bioctahedral and biicosahedral ) give new points @xcite in the moduli space . whether the full moduli space of unitary conformal field theories with @xmath8 has now been accounted for is still to be proven
. however partial results towards this end have been reached @xcite .
the conjectured moduli space for @xmath8 theories is shown in figure [ fig : c=1 ] .
interesting points have been labeled the reader is again referred to @xcite for more details .
the two generalizations of the real line we have considered thus far were that of @xmath26 and that of a circle .
the next case is to combine these to get a torus @xmath59 .
this will give us a conformal field theory of central charge @xmath29 .
again , we will not provide any detailed proofs here but only state the results .
narain @xcite has shown that a string on an @xmath51-torus may be specified by a lattice in a @xmath60-dimensional space of indefinite signature @xmath61 .
the generators of the lattice must have inner product in the form of the matrix @xmath62 thus the question of the the moduli space of conformal field theories is reduced to that of finding the moduli space of lattices in the required form .
the group @xmath63 acts on any lattice to preserve ( [ eq : offh ] ) and this may be taken as the initial guess at the moduli space .
we need to identify points in the space @xmath63 corresponding to identical theories which requires us to take a quotient of this group .
firstly we need to divide by @xmath64 .
these rotations may be viewed as rotations of the @xmath51-dimensional lattice defining the torus itself ( or the set of allowed winding modes ) and the rotations of the dual lattice ( giving the set of allowed momentum modes ) .
these rotations have no effect on the underlying field theory .
thus our candidate moduli space becomes the ( right ) coset @xmath65 ( where the group @xmath64 acts on @xmath63 from the right ) . to complete the picture of the moduli space some more identifications are required .
these come from elements of @xmath63 which act as automorphisms of the lattice .
it is easy to see that such automorphisms must consist of @xmath66 , i.e. , matrices in @xmath63 , preserving ( [ eq : offh ] ) , which have purely integer entries .
assuming we have accounted for all the necessary identifications our moduli space can then be written as @xmath67 where the infinite discrete group @xmath66 acts on @xmath63 from the left .
it is an instructive exercise to interpret this space from the point of view of of the world - sheet action @xcite . in the case of the circle
we could account for the one degree of freedom in the moduli space as corresponding to the radius of the circle .
the natural generalization here would be to put a constant metric @xmath68 on the torus .
thus accounts for @xmath69 dimensions but our moduli space ( [ eq : mod - t ] ) has @xmath70 dimensions . to account for these extra terms
one may add the term @xmath71 to the action , where @xmath72 and @xmath73 indices label coordinates on the world - sheet and @xmath74 is a constant matrix which may be taken to be antisymmetric .
note the factor of @xmath23 in this term this is because of the transformation properties of @xmath75 under the wick rotation from a minkowski signature world - sheet to the euclidean signature we assume in these lectures .
written in terms of complex coordinates on the world - sheet our action then becomes @xmath76 it turns out that in more general cases it is also natural to include the antisymmetric @xmath77-term degree of freedom in the field theory . in order to examine more complicated conformal field theory moduli spaces
it will be necessary to impose further structure .
the structure we will impose shall be supersymmetry .
thus , from this point on , all conformal field theories considered will be superconformal field theories .
the with target space given by a circle can be extended in a straight - forward manner to a superconformal field theory .
to do so we add one free left - moving and one free right - moving majorana - weyl fermion to our action .
such a free fermion is well - known to correspond to the ising model and contributes @xmath78 to the central charge .
our resulting theory will therefore have @xmath79 .
the moduli space of this theory is at first sight little more complicated than the non - supersymmetric case .
the only marginal operator in a generic theory is the one that we already had acting to change the radius of the circle . this should not be surprising since there are no deformations of the free fermion part of the theory .
the moduli space for these @xmath79 theories contains figure [ fig : c=1 ] but it actually has more branches .
the reader is referred to @xcite where this calculation was first performed for more details .
the extra branches in the moduli space arise because one may put different boundary conditions on the fermions .
this moduli space of superconformal field theories with @xmath79 is certainly simpler than the space of all conformal field theories with @xmath79 so adding supersymmetry has simplified the problem .
unfortunately this simplification is still not sufficient to study the moduli spaces of non - trivial examples .
the main topic of these lectures will be conformal field theories with @xmath80 world - sheet supersymmetry .
we insist on @xmath9=2 superconformal invariance in the both sectors and so such theories are often referred to as @xmath9=(2,2 ) superconformal field theories .
it is this extended supersymmetry structure which allows for a dramatic simplification of the problem .
the fundamental reason for this is that in this case one works exclusively with objects taking values in the complex numbers @xmath81 rather than @xmath22 .
holomorphicity then allows many problems to be solved .
to begin our discussion of @xmath9=2 theories let us try to build the simplest theory with a non - trivial moduli space in analogy with the @xmath8 theories and the @xmath82 theories above . adding a second free moving fermion to each sector forces us to add another free boson to complete the @xmath80 supermultiplet .
it appears that our simplest model will have @xmath83 .
actually the moduli space of such theories turns out to be very messy .
extra branches like those that appeared in the @xmath79 theories of @xcite proliferate in this case . to bring this situation under control
let us impose a further condition on our theories .
each field in our @xmath9=2 superconformal field theory has charges @xmath84 under the left - moving and right - moving @xmath3 currents implicit in the @xmath9=2 algebra .
we will insist that all ns fields in our theory will have @xmath85 .
such a constraint appears naturally in string theory it is required for `` spectral flow '' to be a symmetry for fields appearing in the partition function .
the reader is referred to @xcite for more details .
let us compute the complete moduli space for such conformal field theories with @xmath83 .
the following argument was first presented in @xcite and we present here only an outline . for the time being we will assume @xmath86 but this will change later . using the usual notation @xmath19 , @xmath87 and @xmath88 for the fields generating the @xmath9=2 algebra in the left sector
, we have @xmath89 this implies there exists some free boson @xmath41 such that @xmath90 furthermore , we can define fields , @xmath91 , which are neutral under this @xmath3 current : @xmath92 it follows that @xmath93 this in turn implies the existence of a complex free boson @xmath94 such that @xmath95 the free boson @xmath41 by itself forms a @xmath8 conformal field theory as we described earlier . precisely which conformal field it corresponds to is given by the periodicity of the field @xmath41
i.e. , the radius of the circle on which @xmath41 lives
. this is fixed by our quantization condition on the fields in our theory .
a field of charge @xmath96 will contain a @xmath97 factor .
the existence of charge one fields implies that the radius is 1 . from figure
[ fig : c=1 ] we see that this corresponds to a dirac fermion .
that is , we have two fermions with matching boundary conditions .
call this a complex fermion @xmath98 .
we then have that @xmath99 all our degrees of freedom in the moduli space appear to be encapsulated in the field @xmath94 . the first possibility to consider
is that this corresponds to some two - dimensional torus .
the moduli space of such @xmath100 conformal field theories was studied in @xcite .
actually we already know from we what have said that it should be @xmath101 neglecting factors of @xmath55 , one knows that @xmath102 ^ 2.\ ] ] we may map @xmath103 into the upper half complex plane by @xmath104 thus , modulo issues of discrete group actions our moduli space is two copies of the upper - half complex plane .
this space is referred to as the `` teichmller space '' .
the actual moduli space is the teichmller space divided by some discrete group action .
the group @xmath105 contains @xmath106 .
these @xmath107 groups act on the two upper half planes by @xmath108 thus in our moduli space each upper half plane is divided by this modular group . to picture the moduli space we should therefore consider one fundamental domain for this group for each upper - half plane . in our haste in building the moduli space
thus far we have neglected a few @xmath55 factors .
the correct inclusion of these divides the space by two further @xmath55-actions ( see @xcite for this calculation ) .
let us label the two half planes by @xmath109 . as well as the modular group action ( [ eq : modgroup ] ) on each of these parameters we also have @xmath110 the resulting moduli space is shown in figure [ fig : mod - t ] .
when the moduli space of @xmath100 theories was investigated in @xcite it was noted that figure [ fig : mod - t ] was not the complete moduli space . in general
one could take orbifolds of the 2-torus to obtain other conformal field theories with @xmath100 . in the context of our @xmath83 superconformal field theory
any action of a group on this torus which is not simply a translation would act non - trivially on @xmath87 .
thus dividing by such a group would ruin the @xmath9=2 superconformal invariance .
one concludes therefore that the moduli space as shown in figure [ fig : mod - t ] includes all valid @xmath111 superconformal field theories .
it will prove extremely useful to analyze figure [ fig : mod - t ] from the point of view of complex geometry .
let us first note that a 2-torus is a complex manifold of one dimension .
now rewrite the action ( [ eq : torsm ] ) in terms of complex coordinates @xmath25 on the target space : @xmath112 in terms of the action ( [ eq : sm ] ) we appear to have the following degrees of freedom : complex structure on @xmath7 and a choice of @xmath113 and @xmath114 .
can we interpret the moduli space of figure [ fig : mod - t ] in terms of these ?
the complex structure part is easy .
it is well - known that the moduli space of a torus is given by the upper half plane divided by @xmath107 .
we may thus take @xmath115 to represent the complex structure in the usual way , i.e. , the complex structure is that of a torus constructed from the complex plane by dividing by the translations given by 1 and @xmath115 where @xmath116 . it may be shown @xcite that the complex parameter @xmath117 may be built from the remaining degrees of freedom : @xmath118 the real numbers @xmath119 and @xmath120 describe the degrees of freedom of the metric and the @xmath77-term as follows . introduce a constant 2-form @xmath121 such that @xmath122 .
thus @xmath121 generates @xmath123 .
now write @xmath124 where @xmath125 from now on , to avoid cumbersome factors , let us set units so that @xmath126 .
the element in @xmath107 which takes @xmath127 is easy to explain as follows .
we may rewrite the action ( [ eq : sm ] ) as @xmath128 where @xmath129 does not depend on @xmath77 and @xmath130 denotes the pull - back of the two - form , @xmath77 , from @xmath7 to the world - sheet @xmath14 via the map @xmath131 .
first note that this term depending on @xmath77 is `` topological '' it depends only on the cohomology class of @xmath77 .
thus deformations of @xmath77 do not affect @xmath44 .
this means that the equations of motion do not depend on @xmath77 . at the quantum level
@xmath77 is important but correlation functions can only `` see '' @xmath77 via the expression @xmath132 in the path integral .
thus , if we shift @xmath77 by an element of @xmath123 then @xmath44 shifts by @xmath133 for @xmath134 and thus the correlation functions are invariant .
this will be a general symmetry of any theory based on the action ( [ eq : sm ] ) . in the case of the torus it corresponds to @xmath127 .
the other generator of @xmath107 not yet accounted for takes @xmath135 .
if we take @xmath77 to be zero , this amounts to inverting @xmath120 .
since @xmath120 corresponds to the area of the torus , this action amounts to the duality acting on the torus .
the @xmath55-action @xmath136 also has a straight - forward explanation
. one can show that it amounts to taking the complex conjugate of the field theory . in target space terms
this amounts to taking the conjugate complex structure on the target space and changing the sign of @xmath77 .
the last symmetry , generated by @xmath137 , has no classical explanation .
it is `` mirror symmetry '' ( see @xcite for an account of the evolution of this phenomenon ) .
for some points in this moduli space we may also view this as coming from duality .
this occurs as follows .
let the torus be generated by dividing the complex plane by translations by @xmath138 and @xmath139 in orthogonal directions and let @xmath77 be zero .
this gives @xmath140 the duality acting on @xmath141 gives the mirror map @xmath137 .
it is important to realize however that this is a very special case we will not expect there to be any relationship between mirror symmetry and dualities in general .
note that prior knowledge of the existence of mirror symmetry would have been a powerful tool in building this moduli space .
having built the teichmller space we could have generated the whole modular group from the following : 1 .
the modular group for the classical moduli space of complex structures , i.e. , @xmath107 .
2 . the @xmath55 generated by the mirror map .
3 . the @xmath55 generated by complex conjugation plus change in sign of @xmath77 . now let us discuss some of the philosophy behind this latter approach to building the modular group and thus moduli space and contrast it to the method that we used formerly in this section .
the first thing to note is that this latter approach differs markedly from the approach one might consider more natural in physics . as physicists we are used to analyzing space - time in terms of a riemannian metric .
thus when a theory is presented in the form ( [ eq : sm ] ) and one is asked the question `` what is the moduli space for such theories ? '' , one s first reaction will be to find the moduli space of allowed metrics @xmath113 . in the cases
presented thus far in these lectures , the metric is constant and the theory can be solved . indeed
this is the method by which we effectively found the moduli space .
as we shall see , when the metric is not so trivial , life becomes considerably more difficult .
the stunning feature of the program for building the moduli space using mirror symmetry as outlined in the three steps above is that nowhere did we mention the metric ! one might at first think that the complex structure encodes at least part of the metric data since one may think of it as being used to go from the real metric to the hermitian metric used in ( [ eq : sm ] ) . actually this need not be the case as we now explain . the branch of geometry usually used by physicists is differential geometry .
this is implicit when one uses riemannian geometry .
given a manifold @xmath7 , one looks very closely at a very small part of it and assumes it looks roughly like flat space . by
gluing many such small parts together and defining connections one can build a global picture .
there is another method however namely _ algebraic _ geometry . in algebraic geometry ( over @xmath81 )
one considers recovering information concerning the geometry of a space by considering holomorphic functions on that space .
equivalently one may build a space as the zero locus of some function(s ) defined on a simpler space ( such as a complex projective space ) .
one need never mention the metric when studying algebraic geometry .
indeed , one need not insist that the space in question is a manifold and thus it may be difficult to define a metric on our space anyway . the important point to notice
is that the moduli space of complex structures of a space may be determined using the methods of algebraic geometry .
the situation in @xmath9=2 theories is actually twice as algebraic as classical geometry thanks to mirror symmetry . in most cases ( but
not always as we see in section [ s : k3 t ] ) the moduli space roughly factorizes into two parts .
one part describes the complex structure and the other half gives the khler form ( i.e. , hermitian metric ) and @xmath77 .
classically the first half can be determined by algebraic geometry whereas the second half is not such a naturally algebraic object . in string theory
the mirror map relates the second half to a complex structure calculation allowing both halves to be calculated using algebraic geometry ! certainly within the context of @xmath9=2 theories
it would seem unfortunate that physicists have grown to be biased in favour of differential rather than algebraic geometry for analyzing problems .
in fact it is the case that not only is algebraic geometry the more convenient setting for analyzing such string theories but that differential geometry can lead to incorrect conclusions as we shall see later .
this should not really be a surprise .
the fundamental assumption in differential geometry is that if we look closely enough at a small piece of space then it looks like flat space
. why should this be the case in string theory ?
let us now reanalyze the moduli space of the torus from an algebraic point of view .
this may appear to be unnecessarily cumbersome compared to the analytical approach used earlier in this section .
however , since we make no reference to metrics but rather use mirror symmetry this method will be easier to generalize to more complicated examples .
consider the complex projective space @xmath142 with homogeneous coordinates @xmath143 $ ] .
define the hypersurface @xmath7 by the equation @xmath144 an object with one complex dimension defined algebraically is called an `` algebraic curve '' .
algebraic geometers tend to name things as if complex dimensions were real dimensions thus the term `` surface '' refers to two complex dimensions .
the term `` algebraic variety '' is used for the general case of any dimension .
an algebraic curve is a riemann surface and so the topology is defined by the genus .
a curve of genus zero is called `` rational '' and a curve of genus one is called `` elliptic '' . a straight - forward euler characteristic calculation
( see , for example , @xcite ) shows that ( [ eq : ell ] ) defines an elliptic curve .
thus we have an algebraic description of our torus .
as we vary @xmath98 in ( [ eq : ell ] ) we vary the complex structure of @xmath7 . in general
there is a rather subtle relationship between deformations of the polynomial(s ) defining the algebraic variety and deformations of complex structure .
see @xcite for a thorough treatment of this question .
in many simple cases however each nontrivial deformation of the defining polynomial ( i.e. , deformations which can not be undone by a linear redefinition of the coordinates ) define a deformation of complex structure and all deformations of complex structure are obtained this way .
this is the case here for the elliptic curve .
note that linear changes of coordinates @xmath145 and @xmath146 , where @xmath147 is a on - trivial cube root of unity , induce the following transformations of @xmath98 : @xmath148 the moduli space for complex structures of an elliptic curve may be parametrized by @xmath98 modulo the transformations ( [ eq : ellmod ] ) .
note that something special happens at @xmath149 . in this case
@xmath150 factorizes : @xmath151 thus @xmath152 has three components given by the vanishing of each factor .
each factor is a hypersurface defined by a linear equation in @xmath142 which is a rational curve .
thus @xmath7 consists of three rational curves .
each pair of rational curves intersect at one point making a total of three points of intersection .
this space clearly is not a manifold . in general singularities
will occur along a hypersurface defined by @xmath150 when there is a solution to the equations @xmath153 in this case we have solutions when @xmath154 .
these four solutions are mapped to each other by ( [ eq : ellmod ] ) .
this value of @xmath98 is not really allowed in the moduli space for the elliptic curve since it does not give a smooth elliptic curve . for many purposes
however it is useful to add it in to form a compactified moduli space .
let us now relate @xmath98 to the other parameterization of the complex structure using @xmath115 from the previous section .
consider two one - cycles @xmath155 on @xmath7 that generate @xmath156 .
if we cut @xmath7 along these cycles we obtain a parallelogram that can be put on the complex plane .
let us use @xmath136 to denote the complex number parametrized by this plane .
if we rescale by a complex number so that one of the edges lies along the line from 0 to 1 , the other edge will lie along 0 to @xmath115 .
we may also insist that @xmath157 ( since if this fails , simply exchange the cycles ) .
thus , if we define @xmath158 as a holomorphic 1-form on @xmath7 where @xmath150 is a constant we have @xmath159 that is , @xmath115 is defined by the ratio of two `` periods '' .
it turns out that such periods satisfy a differential equation known as the picard - fuchs equation . in general
it is rather cumbersome to set up the machinery for analyzing these differential equations and solutions so here we present only a quick outline of one of the many methods of derivation of the periods .
first let us find a representative for @xmath160 .
in an affine patch where we put @xmath161 and use @xmath162 and @xmath163 as affine coordinates we let @xmath164 this is clearly a holomorphic 1-form .
one may also show that it is everywhere finite by going to the other patches .
thus it may be used to represent @xmath160 .
we now follow @xcite in finding the periods . consider the 1-cycle @xmath165 in the elliptic curve defined by @xmath161 , @xmath166 and @xmath163 by the unique solution to ( [ eq : ell ] ) such that @xmath167 as @xmath168 . here
, @xmath169 is a small positive real number .
this 1-cycle is enclosed by the 2-cycle @xmath170 defined by @xmath171 defined in the ambient @xmath142
. we may thus use cauchy s theorem to give @xmath172 we have neglected overall constant factors at each stage since these are irrelevant . the last step in ( [ eq : pers ] ) is a somewhat formal manipulation to make things more symmetric .
@xmath173 is the 3-cycle in @xmath174 defined by @xmath175 .
let us now rewrite the defining equation in a more general form which allows us to lift questions into @xmath174 : @xmath176 it is now a simple matter to show that @xmath177 actually this last differential equation did not depend on our choice of cycle and will be satisfied by any of the cycles
. thus all periods @xmath178 may be obtained this way .
careful analysis @xcite of the relationship between this affine point of view and our desired interpretation in the projective space shows that @xmath179 , where @xmath180 for some function @xmath150 . for any period on our elliptic curve
we thus obtain @xmath181 this equation will have three independent solutions although we expect only two , since we only have two linearly independent 1-cycles .
the extra solution arises because we did the analysis in @xmath174 .
the extra solution may be removed by omitting the initial @xmath182 in ( [ eq : bigpf ] ) .
( this may be shown by analyzing the monodromy of the extra solution . )
the reader is referred to @xcite for an alternative derivation .
the resulting second order ode is the picard - fuchs equation and is a hypergeometric differential equation .
a general solution is of the form @xmath183 where @xmath184 the value @xmath185 corresponds to our singular elliptic curve and so to map to the fundamental domain as shown in figure [ fig : mod - t ] we map this to @xmath186 .
this point may be considered the fixed point of the @xmath187 symmetry . to recover this symmetry when expressing @xmath115 as a ratio of periods one
is forced to choose @xmath188 for some constant @xmath189 . in order to determine
@xmath189 we need to be able to identify another point in our moduli space . for @xmath190
the elliptic curve admits a @xmath191 symmetry generated by @xmath192 .
this action has three fixed points : @xmath193,[0,-\omega,1],[0,-\omega^2,1]$ ] .
the only torus that admits a @xmath191 symmetry with fixed points occurs for @xmath194 . in this case the symmetry is generated by @xmath195 . therefore we see that @xmath190 is equivalent to @xmath194 .
unfortunately the series ( [ eq : fab ] ) fail to converge for @xmath196 . in order to extend to this region
we need to analytically continue these functions .
this is done by representing the functions as barnes integrals : @xmath197 where @xmath198 is the contour running from @xmath199 to @xmath200 along @xmath201 for some positive real number @xmath202 . closing the contour to the right recovers ( [ eq : fab ] ) .
enclosing to the left recovers different series valid for @xmath196 which are the analytic continuations of ( [ eq : fab ] ) . taking the limit
@xmath203 determines @xmath204 . to complete our moduli space for the @xmath9=2 theories we copy the above structure
we introduce a new algebraic parameter @xmath205 and set @xmath206 as the mirror of our complex parameter @xmath115 .
mirror symmetry now exchanges @xmath205 and @xmath43 and hence @xmath117 and @xmath115 .
simultaneous complex conjugation of @xmath205 and @xmath43 sends @xmath207 and @xmath208 as desired .
we have thus built a complete description of the moduli space without requiring a metric on the target space .
let us consider a target space that is not flat
. that is , we allow @xmath113 and @xmath114 in ( [ eq : sm ] ) to vary over @xmath7 .
it is not possible , in general , to solve this model exactly .
this model is known as the non - linear .
one way of analyzing this model is to look at the @xmath73-functions in perturbation theory @xcite . for conformal
invariance these beta functions must vanish .
suppose @xmath31 is some characteristic radius of the space @xmath7 is some vague sense .
one will expect the perturbation theory to be an expansion roughly in @xmath209 .
thus this method should be reliable in the `` large radius '' limit .
a simple solution to the vanishing of the @xmath73-functions at one loop is given by @xcite @xmath210 and we demand that the metric @xmath113 be khler .
that is , it may be written in terms of a closed ( 1,1)-form @xmath211 given by ( [ eq : bjdef ] ) .
a manifold which admits a ricci - flat khler metric is called a `` manifold . ''
the first condition @xmath212 may be thought of as an assumption from which the conditions follow . as an alternative one might introduce a non - trivial field @xmath213 , indeed one may allow @xmath94 to be a cohomologically non - trivial 3-form so that @xmath77 is only locally defined .
wzw models on a group manifold @xcite are an example of this .
we will not concern ourselves with such models in these lectures mainly because they are very difficult to analyze for anything but the simplest metric .
one might also hope that any such model is equivalent to some model ( if the class of spaces is generalized in some way ) .
therefore , for the purposes of these lectures , if an @xmath9=2 theory has any large radius interpretation then it must be a space .
the constraint of ricci - flatness is actually topological in nature . the first chern class of the tangent bundle on a manifold , which is an element of @xmath123 , is defined in terms of the ricci curvature and so must be trivial for a space . in these lectures we will denote this condition by @xmath214 .
denotes the canonical divisor .
this divisor is associated to the first chern class of the canonical bundle which is turn is equal to the first chern class of the cotangent bundle . ] actually , due to a theorem of yau @xcite , this argument works in reverse .
that is , given a manifold with @xmath214 one may prove the existence of a ricci - flat khler metric . to be more specific ,
given a manifold ( with a given complex structure ) and a suitable cohomology class in @xmath215 there is a _
unique _ ricci - flat metric such that the khler form is a representative of the cohomology class .
thus rather than concerning ourselves with details of the metric one can specify the cohomology class of @xmath211 and let yau s theorem handle the rest . on the face of it , the non - linear with target space @xmath7 , has the following degrees of freedom . 1 . the complex structure of @xmath7 .
2 . the cohomology class of the khler form , @xmath211 .
3 . the cohomology class of @xmath77 modulo @xmath123 . let us use @xmath2 to denote the complex dimension of the target space .
for @xmath4 the space is the torus and we studied that in the last section . for @xmath216
there are two possibilities .
firstly there is the two complex dimensional torus again a flat space .
secondly one might have a `` k3 surface '' . before moving on to @xmath217
we will briefly analyze some aspects of this k3 case . although this is not entirely relevant for the case @xmath217 it will allow us to introduce orbifolds which will be of some importance later on . in the case
@xmath4 we found that the superconformal algebra provided a very restricting set of conditions for the theory .
it was this that forced the target space to be a torus .
for @xmath218 we lose some of the power of this method but for @xmath216 one may still impose certain restrictions . in this case
we can define the boson @xmath219 by @xmath220 we also have the fields @xmath221 which are present in any non - trivial theory with ns fields with integer @xmath3 charges .
the fields @xmath222 together form an @xmath46 affine algebra @xcite .
thus the @xmath3 of the @xmath9=2 superconformal algebra has been elevated to @xmath46 .
in fact , the two fields @xmath87 may be split up into 4 fields transforming as a @xmath223 representation of @xmath46 . in this way
, the algebra is extended to an @xmath9=4 superconformal algebra .
that is to say , in the case @xmath216 , any @xmath9=2 superconformal field theory with ns fields with integer @xmath3 charges is automatically an @xmath9=4 superconformal field theory .
this @xmath9=4 supersymmetry has a striking effect on the perturbation theory for the non - linear .
one may show @xcite that there are no corrections to ( [ eq : cy ] ) at any loop order , nor indeed any nonperturbative corrections .
one may also show that the target space has a quaternionic structure as well as a complex structure .
now let us turn our attention to k3 itself .
what exactly is a k3 surface ?
one way of building one is in the form of an orbifold .
consider the two complex dimensional torus constructed by taking a quotient of @xmath224 , with coordinates @xmath225 , by the translations @xmath226 .
in other words we take the product of two tori with @xmath227 as defined in the last section . the group @xmath55 generated by @xmath228 fixes the 16 points @xmath229 .
this means that when we build the space that is the quotient of the complex two - torus by this @xmath55 we will have a space with 16 singularities .
each of the 16 singularities looks locally like @xmath230 where the @xmath55 in the denominator is generated by @xmath231 .
this latter space has an isolated singularity at the origin .
a space with quotient singularities is known as an orbifold . because such spaces are not manifolds , nave application of differential geometry would be inappropriate .
they are simple to deal with in terms of algebraic geometry however . to see how such objects appear in our moduli spaces we need to introduce the concept of a `` blow - up '' .
to begin with let us define the space @xmath232,(x , y)\in\p^1\times \c^2;\,ay = bx\}.\ ] ] clearly this is a two - dimensional space ( there is one constraint in a three dimensional space ) .
if @xmath233 then @xmath234 is determined and we fix a point on @xmath235 .
thus , away from the origin of @xmath224 there is a one - to - one map between @xmath224 and @xmath236 . at the origin of @xmath224
there is no constraint on @xmath237 $ ] so we recover the whole @xmath235 .
thus @xmath236 looks like @xmath224 with the origin removed and replaced by @xmath235 .
one may also check that @xmath236 is smooth from ( [ eq : smooth ] ) .
the space @xmath236 is said to be obtained from @xmath224 by `` blowing - up '' the origin . given a space ,
one may produce an infinite set of spaces by blowing - up smooth points .
this might at first appear an alarming prospect from the point of view of trying to classify spaces but it turns out that if we take a space and blow - up a smooth point then the resulting space has @xmath238 .
that is , the resulting space will not be ( although it is still khler ) .
the usefulness of the blowing - up construction in the context of spaces stems from what happens when we try to blow - up singular points as we now see .
let us return the the space @xmath230 .
let us define @xmath239 clearly @xmath240 are invariant under the @xmath55 action .
we also have @xmath241 .
in fact this defines a one - to - one map between the space @xmath230 and the hypersurface @xmath241 in @xmath174 .
the latter forms the description of the quotient singularity in algebraic geometry .
let us pretend that @xmath242 $ ] are the homogeneous coordinates of @xmath142 rather than the affine coordinates of @xmath174 .
in this case @xmath241 defines a smooth @xmath243 .
putting coordinates @xmath237 $ ] on this @xmath235 we may map it into @xmath142 via @xmath244 .
the definition of @xmath236 amounted to taking the subspace of @xmath245 where the affine coordinates on @xmath224 represented the homogeneous coordinates on @xmath235 .
let us now play the same game with our subspace of @xmath174 .
that is we consider @xmath246,(x , y , z)\in \p^1\times\c^3;\,xy = z^2,a^2z = abx , a^2y = b^2x , aby = b^2z\}.\ ] ] now , with reasoning similar to before , away from @xmath247 the space looks like @xmath230 whereas at @xmath247 we have a @xmath235 .
that is , we have replaced the singular point of @xmath230 by a @xmath235 and in the process we have ended up with a smooth space
! actually the above form of this space can be simplified .
the constraints are sufficient to uniquely determine @xmath43 from the other variables .
thus the space is isomorphic to @xmath248,(x , y)\in\p^1\times\c^2;a^2y = b^2x\}.\ ] ] in general the notation @xmath249 represents the line bundle over @xmath235 whose first chern class , when integrated over the base , equals @xmath51 .
the condition @xmath214 is only met by @xmath249 when @xmath250 .
thus , it is only when blowing up with @xmath251 that we can hope to obtain a space in two dimensions .
note that blowing - up changes the topology .
in particular , introducing the new @xmath235 adds an element to @xmath252 .
this new element is called an `` exceptional divisor '' .
the term `` divisor '' means , for our purposes , any linear combination of complex codimension one objects in our algebraic variety .
`` exceptional '' refers to the fact that it came from a blow - up .
we may take our space obtained by dividing the complex 2-torus by @xmath55 and obtain a smooth space by blowing up all 16 fixed points .
the hodge numbers @xmath253 can be calculated by starting with the cohomology of the complex 2-torus , projecting out those elements which are not invariant under the @xmath55 action and then adding in the elements from the exceptional divisors of the blow - up .
this is captured by the following hodge diamonds : @xmath254 the diamond on the left is that of the complex 2-torus and on the right that of the desired space . the smooth surface is an example of a k3 surface .
any smooth complex surface which is and is not a torus will be diffeomorphic to a k3 surface ( see for example @xcite ) .
one may divide complex 2-tori by many discrete groups .
so long as this discrete group is a subgroup of @xmath46 one may blow - up the resulting space and , if there were any fixed points , one will find the same hodge diamond as above .
blow - ups of complicated quotient singularities are performed by a sequence of adding @xmath251 spaces @xcite .
let us now consider the khler form on @xmath7 , when @xmath7 is a k3 surface .
then since @xmath255 , the cohomology class of the khler form lives in the vector space @xmath256 .
there is a natural linear mapping between @xmath42-forms and divisors .
this mapping is `` the dual of the dual '' in the following sense .
we take a @xmath42-form @xmath257 and find the dual form @xmath258 such that @xmath259 .
@xmath260 is then taken to be the dual of @xmath261 in the sense that @xmath262 .
let us choose a basis @xmath257 , @xmath263 .
one choice could be to take @xmath264 forming a basis of the @xmath42-forms of the original two - torus and have @xmath265 associated to the 16 exceptional divisors .
we then write @xmath266 using standard khler geometry ( see , for example , @xcite ) one may calculate the area of a curve , @xmath198 , given its homology class , i.e. , area @xmath267 . to calculate this
we need to know the intersection form @xmath268 .
our association between @xmath257 and @xmath260 allows us to rephrase questions in terms of intersection numbers since @xmath269 .
let us consider the intersection numbers of an exceptional divisor , say , @xmath270 , with the other generators .
each exceptional divisor comes from a different fixed point and so need not touch the others .
also a generator of @xmath271 on the original complex two - torus may defined such that it does not pass through the fixed points .
thus @xmath272 for @xmath273 .
what about @xmath274 ?
consider the rational curve , @xmath198 , in @xmath249 defined as the base space while considering @xmath249 as a line bundle .
this bundle may be considered as the normal bundle of the embedding of this rational curve in the space @xmath249 .
deformations of the curve may then be given by holomorphic sections of this normal bundle .
if @xmath275 there are non - zero sections of the bundle with generically @xmath51 zeroes . in this manner , we see that @xmath198 has self - intersection @xmath51 . in our case
@xmath250 and there are no deformations of the rational curve .
however one may assume that the above reasoning extends for @xmath276 and we declare that @xmath277 .
we therefore obtain @xmath278 since it is a reasonable assertion that the area of this curve should be positive , we should therefore impose that @xmath279 .
when one computes all the areas and volumes of all algebraic subspaces of @xmath7 as well as the volume of @xmath7 itself , this positivity condition marks out a region of @xmath256 where the khler is allowed to have values .
the shape marked out is a cone if @xmath211 is a valid khler form then so is @xmath280 for @xmath281 a positive real number .
this subspace of @xmath256 is called the `` khler cone '' .
note that when we take the limit @xmath282 , the area of the exceptional divisor shrinks down to zero .
this is precisely the reverse of blowing - up . by sending @xmath283 for @xmath284
we can thus recover the orbifold of the complex 2-torus . in this sense , the orbifold `` lives '' in the boundary of the khler cone .
this is shown in figure [ fig : kc ] .
it is worth mentioning that one may analyze the orbifold directly in string theory without having to resolve the singularities @xcite . to study the orbifold in @xmath285 , where @xmath286 is a smooth space and @xmath57 is a discrete group acting with fixed points ,
one considers string theory on @xmath286 . as well as
closed strings on @xmath286 one should also consider strings which are open but whose ends are identified under the action of @xmath57 since such strings will be closed on the orbifold .
these are called `` twisted '' strings .
one then projects on to all @xmath57-invariant states .
this procedure allows one to deduce the string spectrum , and hence euler characteristic , of the orbifold @xcite . applying such a method for our two - torus divided by @xmath55 we obtain an euler characteristic of 24 in agreement with the cohomology of k3 .
this calculation is actually more surprising than first meets the eye .
the orbifold itself , despite being singular may still be triangulated and thus singular homology may be defined on it .
the orbifold is a simplicial complex obtained from the two - torus by removing 16 points , dividing by a freely acting @xmath55 and adding back the 16 points .
the euler characteristic is thus 8 .
thus string theory somehow knew that when it was on the orbifold that the associated geometry was that of a ( blown - down ) k3 surface rather than the obvious simplicial complex for purposes of cohomology .
this is another reason to believe that algebraic geometry is the best setting for understanding the geometry of string theory blowing up the orbifold to form k3 is a very natural process in algebraic geometry .
while the correct geometrical picture is now well - understood for simple cases ( see , for example @xcite ) addition of so - called `` discrete torsion '' into the orbifolding process can introduce many curiosities @xcite .
now let us discuss the deformations of complex structure .
deformations of a complex manifold are given by elements of @xmath287 , i.e. , one - forms with values in the holomorphic tangent bundle .
if @xmath7 is a manifold of @xmath2 dimensions , there is a unique @xmath288-form ( up to rescaling ) which may be used to form an isomorphism between @xmath287 and @xmath289 .
see @xcite for a fuller explanation .
it is known @xcite that all elements of this cohomology group give deformations assuming @xmath7 is smooth .
thus the dimension of the moduli space of complex structures is given by the hodge number @xmath290 which in this case is 20 .
it is actually rather tricky to explicitly show all 20 deformations in one model .
let us build k3 as an algebraic variety as we did for the elliptic curve .
we know that any surface we can build that is simply connected will be a k3 surface .
one can show that the hypersurface in @xmath291 defined by an equation of total degree 4 in the homogeneous coordinates @xmath292 $ ] has the required properties ( see chapter 2 of @xcite for a thorough account of such constructions ) .
it is not too difficult to show that writing down the most general quartic defining equation and accounting for linear redefinitions of the coordinates leads to 19 deformations of the defining equation .
thus we appear to `` miss '' one of the the deformations when we analyze the k3 surface in this way .
the reason for this is not too difficult to see .
it comes from the fact that the deformations of complex structure and those of the khler form tend to `` interfere '' with each .
suppose we have an algebraic curve @xmath198 embedded in our k3 surface .
to such a curve we associate a two - form @xmath293 by the linear map discussed earlier .
this two - form will be a ( 1,1)-form by standard results in complex geometry @xcite .
since @xmath198 is an element of @xmath294 , @xmath293 will also be an element of @xmath123 . as we vary the complex structure of @xmath7 , the way that the lattice @xmath123 sits within the hodge decomposition @xmath295 will vary .
thus , the very existence of such a curve @xmath198 will put constraints on the complex structure .
our k3 surface does indeed have such a curve if we define @xmath296 by the vanishing of a generic equation linear in the homogeneous coordinates , then the intersection of this `` hyperplane '' @xmath142 with the quartic hypersurface will be an algebraic curve ( of genus 3 ) .
the fact that there is only class of such curves puts one constraint on the complex structure and lowers the dimension of the moduli space from 20 to 19 .
in fact the general rule is that if we have @xmath51 algebraic curves in the variety generating @xmath51 dimensions of the khler cone , we only have @xmath297 deformations of complex structure .
thus it is hopeless to try to capture the whole moduli space this way .
the actual analysis of the moduli space of k3 surfaces using mirror symmetry is very interesting but we will not pursue it much further here since it is of little relevance to the rest of these lectures .
the reader is referred to @xcite for further details .
what we will present here is the final answer .
it turns out the the moduli space fails completely to factorize into complex structure times khler parts .
the form of the space turns out to be remarkably similar to that of the torus : @xmath298 where the indefinite metric being preserved is of the form @xmath299 , where @xmath300 is the definite even self dual lattice in 8 dimensions and @xmath94 is the 2-dimensional matrix @xmath301 for @xmath302 and @xmath303 for @xmath304 . from narain
s work @xcite this is precisely the moduli space for string on a left - right unsymmetric torus of dimension ( 4,20 ) . why this should be the case is as yet only partially understood @xcite .
note that the moduli space is again of the form of some smooth teichmller space divided out by some modular group .
note also that this moduli space includes orbifolds such as our complex two - torus divided by @xmath55 .
it should be the case ( but this has not yet been checked ) that the modular group should identify at least some of the orbifolds with classically smooth models as predicted in @xcite .
now let us progress on to the case @xmath217 . recall that for @xmath4 there was only one manifold the torus . for @xmath216
we add two possibilities .
how many spaces are there for @xmath217 ? despite considerable effort by mathematicians in recent years , at this point in time it is not known if the number is finite or even if there are any bounds on the euler characteristic !
there are certainly many methods of building manifolds and all constructions which magically produced a k3 surface ( or occasionally a complex two - torus ) for @xmath216 tend to produce something different in each case when applied to @xmath217 .
there are some constructions which can systematically produce a very large number of manifolds .
perhaps the most general construction studied so far is that of complete intersections in toric varieties @xcite .
there are known manifolds which can not be built this way however @xcite .
see @xcite for an introduction to a few of these techniques .
there are some simple statements that can be made .
firstly all spaces with @xmath305 are the complex 3-torus or a k3 surface times a one - torus or some orbifold thereof @xcite .
we will ignore such possibilities here although it should not be difficult using the results of the preceding sections to analyze the moduli spaces completely for these examples .
making the assumption @xmath306 removes some of the complications that occured in the previous section .
the effect of the complex structure on the hodge decomposition @xmath307 will now be trivial since @xmath308 .
this allows us to factorize ( at least locally away from the boundary ) the moduli space into deformations of complex structure and deformations of the khler form .
unfortunately we pay dearly for this simplification .
it was the @xmath309 property of the k3 surface that allowed us to build the @xmath9=4 superconformal algebra that in turn controlled the perturbation theory for the non - linear .
surprisingly one receives no correction to the @xmath73 function for the metric at two and three loop with only @xmath9=2 supersymmetry but at four loops we obtain an extra correction @xcite for the metric : @xmath310+\ldots \label{eq:4lb}\ ] ] thus we no longer have a ricci - flat metric as a solution .
actually this does nt really matter since we never actually found the metric in the last section , we just needed yau s theorem .
we now need some modified version of the theorem to say that given @xmath7 with a complex structure and a point in the khler cone of @xmath7 , there will be a unique solution of ( [ eq:4 lb ] ) ( at least in some large radius limit ) such that the khler form is of the desired cohomology class .
we will assume that this is the case for convenience of argument .
ultimately we do not really care about the metric and this point will be irrelevant .
more worryingly , we now have instanton effects which are non - trivial .
as we shall see this will cause severe problems with a metric - based point of view but fortunately we have the algebrao - geometric methods of section [ s : torus ] to rescue us .
the two factors of the moduli space are best viewed using topological field theory .
we refer to m. bershadsky s lectures in this volume for more details regarding the following summary .
the main point is that a topological field theory has only a few observables and correlation functions which are fairly simple to calculate ( at least compared to non - topological field theories ) .
we can `` twist '' an @xmath9=2 superconformal field theory in two different ways to obtain two distinct topological field theories .
this is achieved via the transformation @xmath311 these two different twistings give rise to the `` a - model '' and the `` b - model '' as defined in @xcite .
only the ( anti)chiral superfields of the @xmath9=2 theory appear in the a and b models .
in particular , only the ( c , c ) and ( a , a ) fields appear in one and the ( c , a ) and ( a , c ) fields appear in the other ( again see @xcite for an explanation of this notation ) .
when analyzed in terms of a non - linear the correlation functions of the fields in the a - model depend only upon the khler form and @xmath77-term of @xmath7 and the correlation functions of the b - model depend only upon the complex structure @xcite .
thus our factors of the moduli space of @xmath9=2 theories coincide with the moduli spaces of the topological field theories obtained by twisting .
thus , although considerable information is lost when one twists an @xmath9=2 theory into a topological field theory , so long as we consider both twists , we do not lose any information about the structure of the moduli space .
let us briefly review each moduli space .
the moduli space of the a - model @xmath312 encodes information concerning the khler form and @xmath77-term of @xmath7 .
this occurs because of instantons .
a solution to the equations of motion of the a - model is a holomorphic map from the world - sheet , @xmath14 , to @xmath7 .
the constant map mapping @xmath14 to a point in @xmath7 is the trivial case .
any other solution is an instanton . if this map is one - to - one then the image of @xmath14 in @xmath7 is an algebraic curve . in the case
@xmath217 it turns out that three - point functions between interesting observables are only non - zero when the world - sheet is genus zero .
the curves in question are therefore rational .
the remaining instantons unaccounted for are the non - trivial many - to - one maps .
these are multiple covers of rational curves ( for which the interested reader is referred to @xcite ) . in the case
that an instanton corresponds to a rational curve , @xmath198 , in @xmath7 , the value of the action is @xmath313 .
let us chose a basis , @xmath314 , for @xmath123 .
using our assumption that @xmath306 we may then make the following definitions @xmath315 it then follows that any correlation function will be a power series in the variables @xmath316 .
note that we can now make more precise our rather vague notion of `` large radius limit '' for the non - linear . as each component of the khler form , @xmath317 , tends to infinity
, we see that @xmath318 .
thus any power series is more likely to converge .
we will define the large radius limit to be the limit in which @xmath319 for all @xmath189 .
thus , not only does @xmath7 become a space of infinite volume but each algebraic subspace will also become infinitely large
. we will consider this large radius limit @xmath320 to be a point in the moduli space .
with this definition we may hope that the correlation functions of the a - model will converge in some non - zero region around this large radius limit point .
the b - model and its moduli space @xmath321 are in many ways much simpler .
the only solution to the equations of motion in this case are the constant maps .
therefore we are free from instantons .
this makes the non - linear very simple to analyze for information concerning observables which appear in the b - model .
essentially the three - point functions calculated at tree - level in the large radius limit will be exact .
actually this fact should have been clear once we noted the factorization of the moduli space into @xmath322 .
if the correlation functions of the b - model depend only on our position with @xmath321 , we can go as close to the large radius limit in @xmath312 as we please thus taking the classical limit .
this has a profound consequence : @xmath323 thus the fact that the left - hand side of figure [ fig : mod - t ] is the moduli space of complex structures for an elliptic curve was not an artifact of the simplicity of this model .
this kind of behaviour persists in higher dimensions . in the case of the complex
one torus we just had to copy @xmath321 to obtain @xmath312 .
in general this situation is not quite so straight - forward .
let us consider a non - linear with target space @xmath7 .
we may then associate to this a conformal field theory and thus a moduli space if @xmath7 is a space .
it may ( or may not ) be the case that there is another space @xmath13 which yields exactly the same conformal field theory as that given by @xmath7 except that we exchange chiral rings ( c , c)@xmath324(a , c ) and ( c , a)@xmath324(a , a ) . in this case @xmath13 is the `` mirror '' of @xmath7 ( again the reader is referred to @xcite . )
if this is the case then clearly @xmath325 and @xmath326 .
for the cases @xmath327 the condition of being a is so constraining that @xmath7 and @xmath13 are topologically the same . thus the mirror map appears as an automorphism on the moduli space .
for @xmath5 it is usually the case that @xmath7 and @xmath13 are topologically distinct ( or even that @xmath13 does not exists as manifold ) .
assuming @xmath13 exists and we can identify it , we may thus calculate the moduli space of @xmath9=2 theories associated to @xmath7 by two complex structure moduli space calculations one for @xmath12 and one for @xmath325 .
the moduli space is then generically @xmath328 where the final @xmath55 quotient corresponds to complex conjugation of @xmath7 and changing the sign of @xmath77 .
before we proceed further to look at a simple example , we first introduce another field theory associated to the target space @xmath7 .
the reason for this short diversion should become clear soon .
this will be witten s linear .
we refer the reader to the original paper @xcite for more details concerning the remainder of this subsection ( see also j. distler s lectures in this volume ) .
the basic idea is that although the conformally invariant for @xmath7 is very complicated we may consider a much simpler with a larger target space containing @xmath7 such that the fields are constrained to live in @xmath7 in some limit .
the latter ( which we will refer to as the _ linear _ ) shares many properties of the nonlinear which has @xmath7 as the genuine target space . before writing down this linear we need to specify our conventions for superspace . to fit in with our earlier description of the @xmath9=2 superalgebra we denote by @xmath329 our coordinates for the left sector and @xmath330 for the right sector .
let us introduce a set of charged ( c , c ) superfields @xmath331 and a set of neutral ( c , a ) superfields @xmath332 in a supersymmetric gauge theory .
that is , @xmath333 where @xmath334 is the gauge covariant superderivative . for our purposes
we consider the case where the gauge group is @xmath335 and @xmath336 .
each @xmath332 may be considered as the supersymmetric version of the field strength for the @xmath3 factor @xcite .
the gauge fields for this theory live in vector superfields @xmath337 .
we may write @xmath338 where @xmath339 is the charge of @xmath331 with respect to the @xmath3 . in this case
@xmath340 is a ( c , c ) superfield with respect to the standard ( rather than gauge covariant ) superderivatives .
we may then expand @xmath340 as @xmath341 note that for a rigid @xmath3 rotation , @xmath342 , the vector superfield is invariant and so @xmath343 .
the details of the vector superfield are rather messy .
the `` @xmath344-term '' appears in the form @xmath345 in the expansion .
a vector superfield in four dimensions has a real vector boson as its lowest component .
when we dimensionally reduce to two dimensions we thus obtain a two - dimensional vector , which we denote @xmath346 and two real scalars which we combine into a complex scalar @xmath347 .
the action to be considered is @xmath348 where @xmath349 is a neutral holomorphic function and @xmath350 and @xmath351 are real numbers .
the equations of motion for this model together with the condition that we lie in the fixed point set of the fermionic symmetries give @xmath352 the solutions to these equations are governed by the topological invariant @xmath353 where @xmath354 is the field strength of @xmath346 .
this is topological since it may be considered to be the integral of the first chern class over a space . assuming single valuedness of sections of the bundle of which @xmath355 is the curvature would force @xmath51 to be integer
this is the two - dimensional analogue of the familiar second chern class @xmath356 term that appears in four - dimensional yang - mills .
one may also show @xcite that @xmath357 .
the first and third equation of ( [ eq : eom ] ) may be combined to complexify the @xmath335 gauge group to @xmath358 @xcite .
( @xmath359 is the group , under multiplication , of nonzero complex numbers . )
we need only work at tree - level for the purposes of these lectures and thus let us assume that @xmath14 is of genus zero . the third equation in ( [ eq : eom ] ) may then be used to tell us that @xmath360 is a holomorphic section of the line bundle @xmath361 .
the only holomorphic section of @xmath362 , where @xmath53 is negative , is the trivial zero section and so we have @xmath363 this will force some , but in general not all , of the fields @xmath360 to vanish .
therefore , generically , the last equation in ( [ eq : eom ] ) forces @xmath347 to vanish . for the classical vacuum
we need to find the classical potential .
this is given by @xcite @xmath364 where the equations of motion set the auxiliary fields as in ( [ eq : eom ] ) .
there is one very useful fact which we should immediately note concerning the parameter @xmath351 .
this appears as the coefficient of the term ( [ eq : c1f ] ) .
@xmath351 behaves just like the `` theta '' in four - dimensional yang - mills theory , which only affects the field theory via instantons and correlations are periodic in this variable . in particular ,
if @xmath51 is an integer then @xmath365
. further discussion of the case @xmath217 and the linear is best illuminated by a specific example . from previous sections we have already discussed two simple methods of construction one might use to build a threefold . the first method from section [ s : k3 t ] would be to take a complex three - torus and divide out by a subgroup of @xmath366 and blow - up the resulting singularities
. the other method would be to take a complex projective space @xmath367 with homogeneous coordinates @xmath368 $ ] and define a hypersurface , @xmath7 , by the vanishing of a generic polynomial of homogeneous degree 5 in @xmath369 . both of these examples appeared in the original paper on manifolds in string theory @xcite .
the latter space above , the `` quintic threefold '' , is particularly attractive because of the simplicity of @xmath312 .
the moduli space @xmath11 of this space was computed first in @xcite .
taking a hyperplane @xmath291 and intersecting with the quintic threefold we obtain a divisor .
( since @xmath217 , divisors will now have complex dimension 2 . )
this divisor generates @xmath370 and thus there is a two - form , @xmath121 , associated ( by the same `` dual of a dual '' construction of section [ s : k3 t ] ) to this divisor that generates @xmath123 .
thus the dimension of @xmath371 is one .
the full hodge diamond of this space is # 1#1 @xmath372 therefore @xmath312 is has complex dimension 1 and @xmath321 has dimension 101 . before discussing @xmath13 and @xmath373
let us explore the linear of @xmath7 .
for our example we set @xmath374 and @xmath375 .
write @xmath376 as @xmath377 and @xmath378 as @xmath379 .
also set @xmath380 where @xmath381 represents other terms ( with arbitrary coefficients ) of degree 5 in the @xmath382 s . for
@xmath349 to be neutral we set the @xmath3 charge of each @xmath382 to be + 1 and the charge of @xmath379 to be @xmath383 .
we will also use the letters @xmath369 and @xmath384 to denote the complex bosons which are the lowest respective components of the associated ( c , c ) superfields in the sense of ( [ eq : spex ] ) .
the classical potential energy becomes @xmath385 ^ 2 + |f(x_i)|^2 + |p|^2\sum_{i=0}^4\left|\frac{\partial f}{\partial x_i } \right|^2 + 2|\sigma|^2\left(\sum_{i=0}^4|x_i|^2 + 25|p|^2\right ) .
\label{eq : thisu}\ ] ] we also have from ( [ eq : eom ] ) that @xmath369 are sections of @xmath249 and that @xmath384 is a section of @xmath386 .
in addition , the fields @xmath360 satisfy the constraint @xmath387 .
suppose for the time being that @xmath388 .
since @xmath389 , we have @xmath390 .
if @xmath391 then @xmath384 must be zero by ( [ eq : eom ] ) .
clearly @xmath392 is non - negative .
let us try to solve for @xmath393 this will be the @xmath394 case . to obtain zero for the first term in ( [ eq : thisu ] ) we must have that at least one of the @xmath369 s is non - vanishing .
the last term then forces @xmath117 to vanish . assuming the transversality condition holds , i.e.
, that ( [ eq : smooth ] ) is only satisfied when all the @xmath369 s are zero , then @xmath384 is forced to vanish by ( [ eq : eom ] ) .
the classical vacuum , i.e. @xmath393 , is thus parametrized by @xmath369 subject to the constraints @xmath395 we also have the @xmath3-action @xmath396 . the first equation in ( [ eq : con1 ] ) gives the sphere @xmath397 .
dividing this by the @xmath3 action gives @xmath398 . the easiest way to see
this latter isomorphism is to build @xmath367 in stages as follows : 1 . take the space @xmath399 with coordinates @xmath400 and remove the origin @xmath401 .
2 . divide any vector in @xmath402 by its length .
this retracts @xmath402 onto the sphere @xmath397 .
it may be considered as the quotient @xmath403 where @xmath35 is the group of positive real numbers .
3 . since @xmath404 and @xmath405 , we have @xmath406 where the @xmath407 acts as @xmath396 .
finally , therefore , our target space consists of the hypersurface @xmath387 in the projective space @xmath367 exactly as desired .
it should be noted that this manipulation extending a @xmath3 quotient to a @xmath359 quotient is essentially equivalent to the procedure referred to earlier for identifying @xmath360 as sections of holomorphic bundles .
deforming the quintic equation @xmath408 will produce deformations of complex structure of @xmath7 .
more interestingly , changing @xmath409 will change the radius of @xmath397 and thus the overall scale of @xmath7 .
this is equivalent to deforming the single parameter governing the khler form of @xmath7 .
thus , @xmath409 appears to be a degree of freedom similar to @xmath211 .
it is also tempting to associate @xmath73 with @xmath77 since they both live on a circle and combine with @xmath409 and @xmath211 respectively to form natural complex parameters .
before we make such a bold identification however we must bare in mind that we have not really built the desired non - linear with target space @xmath7 .
the fields @xmath369 originally span the space @xmath399 .
the metric on this space is the trivial one ( implicitly present in the first term in ( [ eq : linsig ] ) ) .
the metric on @xmath367 is inherited from this original metric by `` symplectic reduction '' @xcite .
this produces the `` fubini - study '' metric on @xmath367 .
the metric on @xmath7 is therefore the restriction of the fubini - study metric .
such a metric is not ricci flat and so can not satisfy ( [ eq:4 lb ] ) in the large radius limit .
actually we have not even the correct degrees of freedom in the linear . in the desired nonlinear
the fields are completely constrained to lie in the target space @xmath7 . while we appear to have the correct behaviour for the low - energy behaviour in the linear one should expect extra massive fields which are not confined to live in @xmath7 .
therefore , in order to obtain the correct degrees of freedom we should integrate out such massive states . such a procedure may well affect the value of the parameters @xmath73 and @xmath409 .
the correlation functions of the a - model depended on @xmath410 .
this dependence came from instanton effects . to compare the quantities @xmath410 and @xmath411 it should be useful to look at instantons in the linear .
the interested reader is referred to @xcite for more features of these instantons .
the instantons are solutions to the equations ( [ eq : eom ] ) but will not necessarily satisfy @xmath393 .
in particular we have that @xmath369 are sections of @xmath249 .
this forces @xmath51 to be an integer and thus @xmath412 as explained in the previous section .
now let us compare this linear instanton with the a - model instantons .
an a - model instanton is a holomorphic map from @xmath14 into @xmath7 .
homogeneity requires @xmath51 to be the same for each @xmath369 . now consider a hyperplane @xmath413 .
the image of @xmath14 under the map given by @xmath369 will intersect this hyperplane @xmath51 times if the map is suitably generic .
this is the _ degree _ of the map . thus far
the linear instantons and the a - model instantons appear identical .
there is a difference however .
for the a - model , the quantities @xmath369 are homogeneous coordinates of @xmath367 and can not simultaneously vanish .
the @xmath369 s of the linear are just sections of @xmath249 and therefore not so constrained .
suppose we have an instanton in the linear where all of the fields @xmath369 vanish at a point @xmath414 and let us assume that @xmath30 .
the equations ( [ eq : eom ] ) and ( [ eq : c1f ] ) dictate that @xmath415 since @xmath388 , the value of @xmath416 must rise rapidly as one moves away from @xmath417 .
in fact , the region on the world sheet around @xmath417 where @xmath416 is not roughly equal to @xmath409 must have an area the order of @xmath418 .
that is , this instanton appears as a small lump around @xmath417 .
we see then that the linear contains all the instantons of a - model and in addition some small - scale instantons which shrink down to points on the world - sheet in the limit @xmath419 . in order to translate between the coordinate @xmath409 and the coordinate @xmath211 on the respective moduli spaces we need to take into account the effects of these small - scale instantons .
in addition to this non - perturbative correction to @xmath409 we should also consider loops in the perturbation theory .
the super - renormalizability of this gauge theory in two dimensions makes such an analysis rather straight - forward .
we need to consider one - loop diagrams such as the tadpole given by @xmath420{\smash{$d_l$ } } } \put(450,615){\makebox(0,0)[lb]{\smash{$\phi_i$ } } } \end{picture } \label{eq : dtad}\ ] ] this results in a shift in @xmath421 which we will denote @xmath422 .
the value of this shift will vary depending upon our position in the moduli space but should be expected to be finite .
let us write @xmath423 and @xmath424 , where @xmath409 is the _ uncorrected _ parameter in the linear ( i.e. , we have not taken into account the one - loop effects ) . to calculate how @xmath425 is related to @xmath426
we combine the one - loop and instanton corrections .
this amounts to @xmath427 where @xmath428 represents the first - order effect from the small - scale instantons and @xmath381 represents the higher orders .
let us introduce a variable @xmath429 as an analogy to the variable @xmath430 of ( [ eq : qdef ] ) : @xmath431 the above sign ambiguity is a problem that always occurs in the problem of trying to find the relationship between @xmath96 and @xmath43 .
the idea is that it appears that one can choose a sign so that @xcite @xmath432 where @xmath433 is a power series in @xmath434 with no constant term .
that is , we assume that the @xmath435 term corrects @xmath426 by 0 or @xmath436
. one can determine this explicitly by counting rational curves in the example being studied as was done in @xcite for example .
this conjecture has yet to be proven in general .
standard renormalization arguments in @xmath9=2 theories guarantee that @xmath433 is a holomorphic function . thus far
the reader may wonder what we have achieved by rephrasing things in terms of the linear .
the answer is that we may now probe the moduli space a long way from the large radius limit at least in some sense the linear may be analyzed away from the limit @xmath419 , i.e. , @xmath437 . as an extreme example
let us consider the case @xmath438 .
now @xmath439 and the fields @xmath369 are forced to be zero ( using the transversality condition when @xmath394 ) . for the classical vacuum
we first look at the vanishing of @xmath392 in ( [ eq : thisu ] ) .
this will force @xmath384 to be nonvanishing .
this in turn forces @xmath117 to vanish .
the equation @xmath387 is now trivially satisfied .
the value of @xmath409 fixes @xmath440 and the @xmath3 gauge symmetry may be used to fix the phase of @xmath384 .
thus all the expectation values of the fields are fixed the classical target space is a point !
actually , to be more precise , we have a landau - ginzburg orbifold theory .
the fields @xmath369 may have a zero vacuum expectation value but their quantum fluctuations are massless and governed by the superpotential @xmath349 which is of degree 5 in the fields @xmath369 .
we also need to note that the field @xmath384 has charge @xmath383 while the charge of the @xmath369 s is 1 . therefore ,
when the phase of @xmath384 was fixed by using the symmetry @xmath441 , we still have a residual @xmath442 symmetry @xmath443 therefore , the fields @xmath369 effectively live in @xmath444 and our theory is an orbifold .
again this theory has instantons .
now @xmath384 will be nontrivial and a section of @xmath386 .
this means that @xmath445 but that @xmath51 need not be integral .
let us consider the instanton given by @xmath446 .
@xmath384 is a section of @xmath447 and will have one zero on @xmath14 .
let us denote this point on the world - sheet by @xmath417 . since @xmath448 we must have that the value of @xmath449 rises quickly to @xmath450 outside a patch of area of order @xmath451 around this zero on the world - sheet .
thus the interesting part of this instanton configuration is confined to a small lump around this single zero .
the fields @xmath369 live in the bundle @xmath452 and will not be single valued .
in particular , the field @xmath384 will pick up a phase @xmath453 with respect to monodromy around @xmath417 whereas @xmath369 will pick up a phase of @xmath454 around the same point @xmath414 .
this is precisely a twist - field configuration where the map @xmath369 is taking the point @xmath417 on the world - sheet to the origin of the target @xmath444 .
thus by varying the value of @xmath409 , we may switch between a target space of a hypersurface in @xmath367 ( for which the only massless modes lie within that space ) and a target space which is a point with massless landau - ginzburg - type massless fluctuations about it .
each of these theories has instantons and in each case the action of the instantons goes as @xmath455 so that the instanton effects become negligible in the large @xmath455 limit .
that is , we have `` exactly '' a theory for @xmath456 ( in an infinitely large target space ) and `` exactly '' a landau - ginzburg orbifold theory for @xmath457 .
we should ask if anything peculiar happens for a finite value of @xmath409 as we change between these two `` phases '' .
the only special value at which something nasty happens occurs classically for @xmath458 . at this value ,
all the fields @xmath369 and @xmath384 may vanish and then @xmath117 may take on any value . any analysis which works for the theory at generic @xmath409 values
should be expected to contain divergences when @xmath409 takes on this value and @xmath117 becomes massless .
actually we have to be a little more careful than this ( and we refer the reader again to @xcite for more detail ) . taking quantum effects properly into account , the minimum energy density for a state at large
@xmath117 goes roughly as @xmath459 .
thus we have a singularity at @xmath460 .
this is the true corrected value for @xmath73 and @xmath409 at the singularity so we need to compute @xmath435 at this point in the moduli space to find @xmath43 .
this may be calculated by assuming that @xmath117 is large and is given purely by the diagram ( [ eq : dtad ] ) .
the result in the case that @xmath374 is that @xcite @xmath461 thus @xmath462 at the singularity .
it turns out that we should choose @xmath463 to get the rational curve count correct @xcite . to summarize
thus far we have identified three special points in the moduli space . at @xmath185
we have a target space of infinite radius .
we also know that in the neighbourhood of @xmath185 we may use ( [ eq : q = z ] ) to relate this coordinate to the coordinate in the moduli space of a - models on the same target space . at @xmath464
the linear is equivalent to a landau - ginzburg orbifold and at @xmath463 the theory is singular .
now let us address the question of @xmath373 which we expect to be isomorphic to @xmath11 .
first let us carefully build @xmath13 .
analysis of the linear showed us that there is a landau - ginzburg orbifold theory in the moduli space .
the same must be true for the theory of a - models ( although at this point in the argument we do nt know where it is located ) and therefore @xmath9=2 conformal field theories . if we go to the correct point in @xmath12 , the landau - ginzburg theory in question may be written as a tensor product of minimal @xmath9=2 models and thus the required landau - ginzburg orbifold is an orbifold of such a tensor product @xcite .
the resulting theory is a `` gepner model '' @xcite .
the process for finding the mirror of a gepner model is well - understood @xcite ( see also @xcite ) .
the result is that the mirror is a certain orbifold of the original gepner model .
this may then be retranslated back into landau - ginzburg orbifold language .
the required result is as follows .
if @xmath465 is the landau - ginzburg theory defined by the superpotential @xmath466 in @xmath444 where @xmath442 is generated by ( [ eq : z5 ] ) , then the mirror theory @xmath467 is the landau - ginzburg orbifold theory with the same superpotential in the space @xmath468 , where @xmath469 is the group consisting of elements @xmath470 where @xmath72 is a non - trivial fifth root of unity and @xmath471 are integers such that the relation @xmath472 holds .
now we need to look at @xmath373 .
this is given by the moduli space of superpotentials .
in the case of @xmath12 there was a 101 dimensional space of superpotentials , or , equivalently , a 101 dimensional space of complex structures on @xmath7 .
the @xmath469 orbifolding group will project most of these deformations out .
as expected , we are left with one parameter .
we may write @xmath150 in the form @xmath473 where @xmath98 is the parameter .
thus , by construction , @xmath190 is the landau - ginzburg orbifold point which is mirror to the landau - ginzburg orbifold point for @xmath7 .
varying @xmath98 will span the space @xmath373 . at this point @xmath13
is a landau - ginzburg orbifold .
as we found with @xmath7 however , the space of @xmath474 will be such that @xmath13 may be pictured as a smooth manifold in part of the moduli space of a - models . since we are only concerned with b - model data for @xmath13 we may take @xmath13 to be this smooth manifold . as a matter of fact
there will be many smooth manifolds which may be used to represent @xmath13 .
this issue will be explained further in section [ s : phase ] .
since we now have a global description of @xmath373 and of linear s on @xmath7 , it is natural to try to map them to each other .
first we should pick good coordinates on each moduli space such that two different values of the coordinate do not correspond to the same point .
the identification @xmath475 in the neighbourhood of @xmath185 makes @xmath43 a good coordinate at least in the vicinity of the large radius limit . around the landau - ginzburg point ,
the instanton number @xmath51 need only be an element of @xmath476 and so it would appear that the identification should be @xmath477 . actually there is a @xmath442 symmetry around the point @xmath464 which identifies theories such that we restore @xmath475 . as indicated earlier , the instantons with non - integer @xmath51 may be thought of as twist fields . a twist field @xmath478
then transforms under this @xmath442 symmetry as @xmath479 therefore @xmath43 is a good coordinate locally around @xmath464 .
picturing the moduli space as @xmath235 with @xmath43 the usual affine coordinate , we have done enough to show that @xmath43 is a good coordinate everywhere ( together with the point @xmath464 ) . the @xmath442 transformation generated by @xmath480
can be made a symmetry of ( [ eq : ydef ] ) by extending it so that @xmath481 .
therefore @xmath98 is not a good coordinate .
this symmetry of @xmath98 is the only one induced by symmetries of the equation ( [ eq : ydef ] ) .
this means that @xmath482 is a good coordinate .
we now wish to map these two @xmath235 moduli spaces to each other in a one to one manner .
the pervading @xmath9=2 structure present means that we expect this map to be holomorphic .
this forces @xmath483 where @xmath484 are complex numbers such that @xmath485 .
we may thus fix this map by identifying three points .
we have already identified @xmath464 and @xmath190 as the landau - ginzburg orbifold point .
we know that @xmath463 corresponds to a singular theory
. the equation ( [ eq : ydef ] ) satisfies ( [ eq : smooth ] ) when @xmath486 .
this therefore provides a natural candidate for the singular point in @xmath373 .
actually , if one attempts to calculate correlation functions using the `` chiral ring '' of @xcite then it is precisely when ( [ eq : smooth ] ) is satisfied that these calculations become badly defined .
thus we map @xmath463 to @xmath486 . for our last point
we pick the large radius point @xmath185
. clearly it would be unnatural for this point to be anything other than @xmath487 . to check this
however one may put the zamolodchikov metric on @xmath11 and @xmath373 .
the point @xmath185 maps to @xmath488 in @xmath11 which corresponds to the large radius limit and is thus an infinite distance away from other points on @xmath235 .
the same is true for @xmath487 on @xmath373 @xcite .
we have thus fixed @xmath489 it is amazing just how simple ( [ eq : zpsi ] ) is .
this is the justification of our using the linear .
while the map between @xmath11 ( in terms of @xmath96 ) and @xmath373 ( in terms of @xmath98 ) is rather complicated , there is a very simple relation between the linear version of @xmath11 ( in terms of @xmath43 ) and @xmath373 . at this point
it is very tempting to end our analysis of the moduli spaces since in many ways we have all the information we need concerning the moduli space .
the only problem is that we have described @xmath11 in terms of the coordinate @xmath43 rather than @xmath96 .
recall that @xmath96 was derived from the khler form and thus ultimately differential geometry .
it appears that @xmath43 is the coordinate that appears more naturally in string theory and perhaps we should rewrite general relativity in terms of degrees of freedom naturally expressed in terms of this parameter rather than @xmath96 .
rather than attempting such an ambitious problem we will submit to the present conventions for describing space - time and try to reparametrize our moduli space in terms of @xmath96 .
actually the analysis of section [ ss : t - alg ] allows us to do this without much more work .
recall that we found @xmath490 in terms of a ratio of periods on the mirror , in this case @xmath13 .
one may use the local geometry of the moduli space @xmath9=2 theories to show that this is also the case for any @xmath2 @xcite .
therefore we just need the picard - fuchs equation for these periods to find @xmath96 as a function of @xmath43 .
the desired equation for the periods @xmath178 is @xcite @xmath491 the four solutions of this hypergeometric differential equation are characterized by their monodromy around @xmath185 , which goes as @xmath492 respectively . the powerful constraint ( [ eq : q = z ] )
then determines the answer uniquely to be @xmath493 where @xmath494,\cr}\ ] ] where @xmath495 is the digamma function defined as the derivative of @xmath496 .
these periods may also be found directly from a smooth model of @xmath13 without going via the picard - fuchs equation @xcite .
we may now expand @xmath96 as @xmath497 which converges for @xmath498 .
thus , by using mirror symmetry we have been able to determine the precise form of ( [ eq : q = z ] ) .
it should come as no surprise that this series has a radius of convergence equal to @xmath499 since the singular theory lies at @xmath463 . at the point
@xmath463 the series just manages to converge and we obtain @xmath500 we denote this value of @xmath120 by @xmath501 . if we have a manifold diffeomorphic to the quintic hypersurface and if the size of it is such that the area of a curve which generates @xmath252 is larger than about @xmath502 then we can place precisely where our model lies in the moduli space of @xmath9=2 theories by using ( [ eq : q = zquin ] ) .
we have not yet accounted for the complete moduli space of these theories however .
what about the region where @xmath503 ? at this point the series fails to converge .
if we try to compute correlation functions using the nonlinear for a point in this region of the moduli space we will also find that the instanton corrections form a divergent series too . if we were to be pragmatic at this point we should say that the nonlinear is not such a good picture for any theory in the region @xmath503 and that some other picture is more appropriate .
indeed , we already know what the other picture is we should interpret such theories as landau - ginzburg orbifolds corrected by instantons . actually , to use more conventional conformal field theory language , these instantons are twist fields .
one may deform a conformal field theory orbifold by a twist field which is also a truly marginal operator . although the resulting theory will not generically be an orbifold
, one may compute correlation functions as a power series in terms of the fields in the original orbifold theory .
this will be a power series in the coupling to the marginal operator in the usual way in conformal perturbation theory .
one will find that such a series will converge when @xmath503 .
for any point ( except when @xmath504 ) in the moduli space @xmath11 we may therefore associate an effective target space theory . in one region
we have a manifold with instanton corrections and in the other region we a landau - ginzburg orbifold perturbed by twisted marginal operators .
both types of theories have correlators in the form of convergent power series .
the regions do not overlap .
this leads to the `` phase '' picture of the moduli space @xcite .
we do not have some universal picture for a target space @xmath7 associated to @xmath11 .
it is necessary to have a set of target space descriptions which are applicable in various regions in @xmath11 .
the attentive reader may have realized that , because we have holomorphic functions , we should be able to analytically continue our series beyond their radius of convergence .
certainly this is possible as we now explain .
one can therefore begin to interpret one phase of the moduli space in terms of other phases .
the more cautious reader might declare such a process to be unnatural however .
anyway , let us see what happens when we extend the picture into the landau - ginzburg region . to analytically continue we first need to apply branch cuts .
the branch points are clearly @xmath505 . in our identification of the strip @xmath506 , where @xmath507 , with the hemisphere @xmath498 we have already cut from @xmath185 to @xmath463 .
we require another cut emanating from @xmath464 . to avoid changing the hemisphere around @xmath185 which is already in the desired form , we cut form @xmath464 to @xmath463 .
this latter cut may be performed by returning to our variable @xmath98 and imposing @xmath508 the same method of barnes integrals as was used in section [ ss : t - alg ] then yields @xmath509 as a convergent power series for @xmath510 .
we may now map the entire @xmath235 of @xmath11 into the @xmath506-plane .
the result is shown in figure [ fig : bjquin ] .
the first striking feature of figure [ fig : bjquin ] is that not all values of @xmath120 are allowed .
the sphere is mapped to the striped region and so @xmath120 acquires a minimum value of @xmath511 .
this fits in with the idea of a minimum allowed size as we saw for the circle .
any conformal field theory in this class may be associated with a manifold of a size scale no less than the order of the planck length .
the shaded area in figure [ fig : bjquin ] represents the area in which the instanton sum for the nonlinear converges . depending on one s taste
, one might wish to declare this to set the minimum length and then say that the bottom region must be describe in terms of landau - ginzburg orbifolds .
a very important point to note is the following .
the striped region in figure [ fig : bjquin ] may at first sight look similar to a fundamental @xmath107 region in the upper half plane as we had for the torus example in section [ s : torus ] .
sadly the region in figure [ fig : bjquin ] _ can not _ be obtained this way .
one may map this region by a transcendental function to another region which may be written in the form of a fundamental region @xcite but as this map is not one - to - one it is not clear what purpose this can serve for a model of the moduli space .
the moduli spaces for the torus and for k3 surfaces appeared naturally in the form of a teichmller space divided by some modular group .
it appears that for more generic @xmath9=2 theories such a description is no longer valid .
this should not dishearten the reader however we were still able to find @xmath11 without such a description .
in the last section we looked at the simplest case of a moduli space for a nontrivial case @xmath217 .
it was found that the space naturally divided into two phases . in this section
we will discuss the more general picture . to facilitate the discussion it will prove useful to run through another example .
this example was first analyzed in @xcite ( see also @xcite ) .
the idea is very similar to that of the quintic hypersurface except this time we begin with a _ weighted _ projective space @xmath512 .
this is the space with homogeneous coordinates @xmath368 $ ] , omitting @xmath513 $ ] , where we identify @xmath514\cong[\lambda^2x_0,\lambda^2x_1 , \lambda^2x_2,\lambda x_3,\lambda x_4],\ ] ] where @xmath515 .
the desired hypersurface is given by @xmath516 this hypersurface is not smooth .
there is a @xmath55 singularity along the surface @xmath517 $ ] in @xmath512 which intersects our hypersurface along a curve . at each point on the curve , this singularity is essentially the same as the the one studied in section [ s : k3 t ] .
thus we may blow each point of the curve up to get @xmath235 to smooth the space .
the resulting exceptional divisor , @xmath518 , will be a surface ( basically the old curve@xmath519 ) .
this smooth manifold , containing the surface @xmath518 , will be our smooth model for @xmath7 .
the hyperplane in @xmath512 produces another divisor in @xmath7 which we call @xmath520 .
the divisors @xmath518 and @xmath520 are associated to two linearly independent elements of @xmath371 .
in fact @xmath521 and these elements form a basis
. therefore @xmath11 is now two - dimensional . for witten s linear
we now want @xmath522 and so we need two @xmath3 charges and on complexification have two @xmath359 actions to consider on the target space .
we assign the following charges : @xmath523 and have the following superpotential @xmath524 where @xmath381 represents other term that may be added ( with arbitrary coefficients ) respecting the quasi - homogeneity of the equation .
the @xmath344-terms are now fixed by the equations of motion as @xmath525 recall that for the example in section [ ss : eg1 ] we found that fields which were sections of @xmath249 vanished when @xmath439 even though one might initially expect this only to be the case when @xmath276 .
the @xmath394 case should be checked by the transversality condition .
we assume this to be the case below except where noted . we obtain four phases : 1 .
@xmath526 and @xmath527 : in this case @xmath528 and @xmath529 by the condition @xmath530 . the fields @xmath531 are sections of @xmath532 and the fields @xmath533 are sections of @xmath534 .
therefore they all vanish .
taking the @xmath344-terms to vanish to solve @xmath393 forces a non - zero value for @xmath535 and @xmath384 which may be fixed by the @xmath536 gauge group .
thus the target space is a point .
there is a residual @xmath537 symmetry left over so that our theory is actually a landau - ginzburg theory with target space @xmath538 .
2 . @xmath526 and @xmath539 : in this case @xmath540 which forces @xmath541 . solving for @xmath393 , the @xmath344-terms force @xmath535 to be non - zero and
that @xmath542 can not simultaneously vanish .
we may use one @xmath359 symmetry to fix @xmath535 .
the other one may be used to turn @xmath542 into the homogeneous coordinates of @xmath512 .
the target space is thus the hypersurface @xmath152 in @xmath512 .
we call this the `` orbifold '' phase since this hypersurface has @xmath55 quotient singularities .
3 . @xmath543 and @xmath544 : now @xmath545 forcing @xmath546 and @xmath529 forcing @xmath547 . solving for @xmath393 forces
@xmath384 to be nonzero and @xmath548 and @xmath549 to not simultaneously vanish .
one @xmath359 may be used to fix @xmath384 and the other to turn @xmath533 into homogeneous coordinates on @xmath235 .
the target space is therefore @xmath235 .
this is not quite the full story however , the fluctuations of the fields @xmath550 are massless and governed by a quartic superpotential .
fixing @xmath384 leaves a residual @xmath551 symmetry meaning that that fields @xmath550 live in the space @xmath552 .
this phase is a hybrid - like phase consisting of a landau - ginzburg ( orbifold ) bundle over @xmath235 .
we call this the `` @xmath235 phase '' .
4 . @xmath543 and @xmath553 : this case is a little more complicated and we have to be more careful concerning our assertion that sections of @xmath362 are forced to vanish when @xmath554 .
we have @xmath540 which will force @xmath541 when @xmath555 and we will find that @xmath541 by transversality when @xmath556 .
when @xmath557 we will have @xmath546 but we need to exercise more care when @xmath558 .
solving for @xmath393 we find that @xmath533 can not simultaneously vanish and that @xmath550 can not simultaneously vanish .
first assume that @xmath559 .
we may then use one of the @xmath359-actions to fix @xmath535 . let the other @xmath359 make @xmath542 into homogeneous coordinates .
our target space is now very similar to that of the hypersurface in @xmath512 except that we are missing the points where @xmath560 and @xmath561 .
now let @xmath546 .
now we have @xmath561 and @xmath548 and @xmath549 may be any value except both zero .
use one of the @xmath359-actions to turn @xmath533 into the homogeneous coordinates of @xmath235 and the other @xmath359-action turn @xmath531 into homogeneous coordinates . what we have is a description of the smooth , blown - up manifold described above the singular point set has been replaced by an exceptional divisor consisting of a curve times @xmath235 .
this is the `` '' phase .
this produces a `` phase diagram '' which we show in figure [ fig : ph ] .
the instantons which appear in each sector can now appear in various forms .
the cases are given as follows : 1 .
for the landau - ginzburg orbifold we have only twist - fields .
2 . for the orbifold phase we have two kinds .
firstly we have rational - curve type instantons ( together with their small - scale supplements ) .
secondly we have twist - field instantons confined to the region around the @xmath55-quotient singularity .
3 . for the @xmath235 phase
we have rational curve type instantons the rational curve being @xmath235 itself !
we also have twist - field instantons in the landau - ginzburg fibre since this has a @xmath551 quotient singularity .
4 . for the phase
we have rational curve instantons .
now let us consider the mirror , @xmath13 , of this example .
this is obtained by dividing @xmath7 by the group @xmath562 consisting of elements @xmath563 where @xmath564 and @xmath565 are all integers .
the general form of the defining equation for @xmath13 is then @xmath566 where @xmath41 and @xmath98 are complex parameters that vary the complex structure of @xmath13 .
this is in agreement with our expectation that @xmath567 . to map the space spanned by @xmath41 and @xmath98 to the moduli space of linear s for @xmath7 we need to find special points in these spaces .
we first consider the values of @xmath98 and @xmath41 for which the hypersurface ( [ eq : fy2 ] ) becomes singular .
a little algebra shows that @xmath568 for all @xmath369 admits nontrivial solution when @xmath569 or @xmath570 . with some foresight
let us introduce the variables @xmath571 we may now map the singular points in our moduli space into the @xmath572 plane .
that is we find the values for @xmath573 and @xmath574 for which there can be a @xmath41 and @xmath98 which give a singular @xmath13 .
the result is shown in figure [ fig : ph2 ] now imagine that we `` zoom out '' infinitely far from figure [ fig : ph2 ] .
it should be easy to see that the shaded region will become precisely the phase boundaries in figure [ fig : ph ] .
in fact , the variables in ( [ eq : rhos ] ) were chosen precisely so that we may make the identification @xmath575 and @xmath576 .
the fact that the asymptotes of the region in figure [ fig : ph2 ] are parallel to , rather than along , lines passing through the origin comes from the one - loop renormalization effects of @xmath350 just as we had a @xmath577 shift in section [ ss : eg1 ] . thus assuming a simple map between the b - model moduli space coordinates and the linear coordinates as we had in ( [ eq : zpsi ] ) , we obtain @xmath578 the reader is referred to @xcite for the complete proof that this is indeed the correct map .
again we may calculate the picard - fuch s equation for this two - parameter example @xcite . again ( [ eq : q = z ] ) uniquely determines the relationships @xmath579 this is a convergent power series in the calabi - yau phase . as before each phase leads to some notion of a series for correlation functions which will be convergent in some region .
in the example studied in section [ ss : eg1 ] we found that any generic point in the moduli space lay in the region of convergence of one of the power series .
this is no longer the case in this two parameter example .
the shaded region in figure [ fig : ph2 ] marks where none of the power series converge . in this sense
, the phase picture does not cover the entire moduli space although it will cover most of a circle with center @xmath580 in the limit of infinite radius .
as we did in section [ ss : eg1 ] we may analytically continue the @xmath430 s into the other three phases .
this calculation was done in @xcite .
doing so one obtains the following bound for the entire moduli space @xmath581 this is a very important result .
it means that the algebraic curves in @xmath7 whose areas are measured by @xmath120 ( i.e. , the rational curves in the exceptional divisor ) may shrink down to zero size .
* we have broken the constraint of a minimum length .
* the reader may argue that we analytically continued @xmath430 beyond the phase and so we may artificially have been able to reach zero size .
actually in this case , @xmath120 can take on arbitrarily small values while one remains in the phase @xcite .
thus one really is forced to reject a notion of minimum length in this example .
it is worth noting that general arguments have been made which appear to prove a universal notion of minimum length ( such as @xcite ) .
these arguments appear to rely on the assumption that space - time is locally flat , i.e. , one believes the differential geometry view . unless we can concoct some alternative view of small distances this will appear to show that differential geometry is seriously misleading at small scales !
the examples studied so far show many of the aspects of the global structure of the moduli space @xmath12 is rather dull and can be explained classically while @xmath11 forces a `` phase '' description upon us and forces us to rethink our notions of geometry at small scales .
let try to understand the more general picture of @xmath11 .
one of the phase transitions for the example studied in section [ ss : eg2 ] was between an orbifold and a manifold .
it was precisely the blow - up of section [ s : k3 t ] .
all of the phase transitions mentioned thus far are actually of this type .
let us return to the example of the quintic threefold of section [ ss : eg1 ] .
the two phases here were a smooth manifold and an landau - ginzburg theory in @xmath444 .
the natural question to ask is if we can blow - up the quotient singularity @xmath444 .
the answer is yes , and the procedure is very similar to that of blowing up the @xmath230 singularity . in the latter case we used the line bundle over @xmath235 with first chern class @xmath582 .
for @xmath444 we use the line bundle over @xmath367 with first chern class @xmath383 .
that is , the exceptional divisor is @xmath367 but @xmath367 is precisely the ambient space for the phase .
this leads us to the following general picture .
define @xmath7 as the critical point set of some function @xmath349 on some ( non - compact ) space , @xmath583 . by blowing up and down quotient singularities in @xmath583
we induce the phase transitions in @xmath7 .
the singularity @xmath538 is blown up using an exceptional divisor with two irreducible components . having each of these components blown up or down leads to the 4 phases .
for example , the first component is @xmath584 .
blowing up the landau - ginzburg phase using this component gives the @xmath235 phase .
this still has @xmath551 singularities all along this @xmath235 .
the second component is @xmath367 .
this component resolves the @xmath551 singularities in each fibre .
the restriction to the critical point set of @xmath349 in each fibre forms a quartic constraint .
a quartic in constraint in @xmath367 is a k3 surface .
thus this latter phase is a fibre bundle over @xmath235 with generically a k3 fibre .
this is one description of the smooth manifold @xcite .
we may also reach the smooth phase by blowing up along @xmath367 first to obtain the orbifold phase and then along @xmath585 to resolve the orbifold .
have we constructed the general picture by considering blow - ups of quotient singularities ?
the answer to this question is no .
we should then ask the question as to whether there is any transformation that might give the general picture .
the answer that appears to be the case is that we consider `` birational '' transformations .
birational transformations occur very naturally in algebraic geometry and thus we should nt be too surprised that they will be natural objects in @xmath9=2 theories .
two algebraic varieties @xmath586 and @xmath587 are birationally equivalent if one can find open subsets @xmath588 and @xmath589 such that the set of functions in @xmath590 is isomorphic to the set of functions in @xmath591 ( see @xcite for a more careful definition ) .
an example of birationally equivalent pairs are given by @xmath586 being a blow - up of @xmath587 .
another example of birational equivalence can be provided still from quotient singularities .
this stems from the fact that the process of blowing - up a quotient singularity need not be a unique process .
suppose we have a singular space @xmath592 which may be smoothed by blowing - up into two topologically distinct smooth spaces @xmath586 and @xmath587 .
these two smooth spaces are then birationally equivalent . in @xcite
an example with @xmath593 was studied .
this requires toric geometry which takes us beyond the scope of these lectures so we will not provide details here .
there are 100 phases in all for this model of which 5 consist of smooth manifolds .
these 5 manifolds are all the possible resolutions of an orbifold which provides one of the other phases
. one may also picture a direct transition between these manifolds in the form of a `` flop '' as we now explain .
we can raise our example of a @xmath230 quotient singularity of section [ s : k3 t ] to one higher dimension by considering the space @xmath594 in @xmath595 with coordinates @xmath596 .
transversality tells us that this hypersurface has an isolated singularity at @xmath597 .
it can not be written as a quotient singularity .
now consider the space @xmath598,(x , y , z , w)\in\p^1\times\c^4;\ , az = bx , ay = bw\}. \label{eq : flop1}\ ] ] this smooths the singularity , in a manner similar to blow - ups discussed earlier , replacing the singular point by @xmath235 .
the space @xmath598,(x , y , z , w)\in\p^1\times\c^4;\ , aw = bx , ay = bz\ } \label{eq : flop2}\ ] ] does pretty much the same thing .
the only difference between ( [ eq : flop1 ] ) and ( [ eq : flop2 ] ) is the way in which the @xmath235 is inserted .
the two smooth spaces produced are said to differ by a flop . this local picture
may be fitted into more complicated global geometries .
these five smooth models for @xmath7 occupy adjacent phases ( since they are related by flops ) in the phase picture and so we may consider a path of conformal field theories passing from one to the other .
so long as this path is generic , it will not hit the singularity lying in the complex codimension one `` phase boundary '' and so this transition is perfectly smooth from the conformal field theory point of view .
this shows how string theory can give rise to a `` smooth topology change '' in the target space so long as the two topological spaces are birationally equivalent as algebraic varieties . if the reader was not convinced thus far that algebraic geometry was superior to differential geometry for our purposes then this last point must surely convince them .
the natural equivalence class in differential geometry is that of diffeomorphic equivalence which is a class stronger than topological equivalence .
string theory happily combines different classes smoothly into the same moduli space string theory is oblivious to such distinctions
! however , every phase in a given moduli space belongs to single birational equivalence class .
there are many more possibilities of birational transformations than those provided by resolving quotient singularities and flops .
one of the more exotic ones is the `` exoflop '' of @xcite . in the case of the flop
, one @xmath235 is transformed into another one within a manifold . in the the case of an exoflop
, a rational curve is taken from within the manifold to one glued onto the outside .
the complete picture of possible phases is far from complete . the most general construction for models for which one understands the linear is that provided by @xcite .
these are explored to some extent @xcite although the full range of possibilities have yet to be classified .
one interesting point discussed in @xcite is that the phase diagrams need not contain any smooth phase .
this really is quite natural given that there is nothing particularly special about the smooth phase it has no more reason to exist than any of the other phases .
the reason we tend to like manifolds rather than the other phases is because we tend to be happier with differential geometry which only works properly in this phase . having said that we should rid our minds of any bias towards the manifold phase we should note some rather interesting curiosities which occur when we try to analyze the whole moduli space in terms of these smooth phases .
let us consider the five parameter example of @xcite which has 100 phases of which 5 are manifolds .
when we do the analytic continuation analysis to map out the entire moduli space in terms of one phase , which phase should we begin with ?
the surprising answer is that is does nt actually matter as we now explain .
suppose we begin with one of the topologies for @xmath7 which we denote @xmath586 .
thus we have some phase in the moduli space containing the large radius limit of @xmath586 where we put @xmath599 .
now let @xmath587 be another topology which is obtained by flopping @xmath586 .
this flop will correspond to a rational curve @xmath600 and another rational curve @xmath601 which arise in the form of equations ( [ eq : flop1 ] ) and ( [ eq : flop2 ] ) .
we may chose @xmath602 such that @xmath603 now we consider the flop process . to make things simple
we should make every rational curve in @xmath586 have infinite area except for @xmath270 .
this may be achieved by setting @xmath604 .
now we analyze the picard - fuch s equation to obtain @xmath602 in terms of @xmath605 .
the result is @xcite @xmath606 analysis of the phase picture shows that if we introduce coordinates @xmath607 for a patch of coordinates with origin at the large radius limit of @xmath587 then @xmath608 .
we also have @xmath609 .
this means that the analytic continuation between these two phases simply asserts that the area of @xmath270 is minus the area of @xmath610 .
thus we naturally identify a topology with a certain conformal field theory by demanding that all areas be positive .
it appears to be the case @xcite that the following now applies in the case that there is at least one phase .
given any conformal field theory we may associate some target space topology ( possibly by analytic continuation ) in which all areas are non - negative .
that is to say when we go from a phase picture in terms of @xmath350 s which covers the whole space @xmath611 and then remap this space into the space of @xmath612 s then only points within ( or on the boundary of ) cones corresponding to smooth phases are covered .
this is shown schematically in figure [ fig : ka ] for a hypothetical example with @xmath522 and five phases @xmath613 for which only @xmath586 and @xmath587 are smooth manifolds .
we hope the reader is convinced that algebraic geometry together with mirror symmetry is a very useful tool for analyzing the moduli space of @xmath9=2 theories .
although sticking to metric - based ideas seems to work without any problems for some models , it appears to face severe short - comings in the generic case . in particular , when there are instantons one tends to have phases since one will have instanton sums which are only convergent in a certain region of the moduli space .
when this happens the other phases appear to have an equal right to be taken as a model for the target space and then one necessarily needs to discuss singular spaces .
where metric - based ideas may work , in the case @xmath217 , appears to be restricted to the few cases where @xmath614 ( which implies that @xmath615 which in turn implies that the target space has a continuous isometry ) .
although we did nt discuss these examples in detail , the interested reader may try to apply the methods of section [ s : torus ] and [ s : k3 t ] for these cases .
the methods used in section [ s : d=3 ] have certain shortcomings which may not have been completely apparent .
firstly there may be some subtleties introduced into the phase picture if there are some symmetries of the defining equation that are not of the most nave kind .
we refer the reader to @xcite for such an example .
another , more severe , problem is that not all of the dimensions of the moduli space may be parametrized by the linear .
that is , we may only be able to write down models based on a gauge group @xmath335 where @xmath616 .
the mirror to this statement is the fact that not all deformations of a complete intersection can , in general , be written as deformations of the defining equations @xcite .
we do not have any techniques at our disposal to address the complete moduli spaces of such objects at this point in time .
we should also repeat that some @xmath9=2 conformal field theories can not be put in the form of a gauged @xmath3 linear at all .
this is equivalent to the statement in geometry that not all manifolds can be written as complete intersections in toric varieties .
one may try to extend methods to nonabelian gauge groups .
although such models were discussed in @xcite , since the mirror map construction for such models is not yet understood , we may not apply most of the methods discussed in this paper .
also there is no reason to believe that these nonabelian groups will exhaust all the @xmath9=2 theories .
although the moduli space of @xmath9=2 theories is far better understood now than people imagined a few years ago it appears that there is still much left to discover .
it is a pleasure to thank b. greene , d. morrison and r. plesser for many useful conversations .
the work of the author is supported by a grant from the national science foundation . | we review the structure of the moduli space of particular @xmath0 superconformal field theories .
we restrict attention to those of particular use in superstring compactification , namely those with central charge @xmath1 for some integer @xmath2 and whose ns fields have integer @xmath3 charge .
the cases @xmath4 , 2 and 3 are analyzed .
it is shown that in the case @xmath5 it is important to use techniques of algebraic geometry rather than rely on metric - based ideas .
the phase structure of these moduli spaces is discussed in some detail .
epsf # 1date_#1 _ # 1thefnmarkfootnotetext1991 _ mathematics subject classification .
_ # 1 # 1thefnmarkfootnotetext _ key words and phrases .
_ # 1 ps .
@firstpageps .
@empty oddhead evenheadoddhead ifundefinedreset@font@font footnotetext#1 # 1 o p # 1#1 # 1 # 1 # 1 # 1 # 1#2 # 1#2_#1f_#2 # 1#1 # 1#1 _ # 1@xmath6 0 1.5 cm |
thermal emission from neutron stars can be used to measure the magnetic field , temperature , and composition of neutron - star envelopes , and to constrain the properties of matter under extreme conditions ( see , e.g. , @xcite , and references therein ) . to achieve these goals
, one should use reliable models of the atmosphere or condensed surface , where the thermal spectrum is formed , and of deeper layers , which provide thermal insulation of hot stellar interiors . in these layers , the effects of strong magnetic fields can be important . in recent years , significant progress has been achieved in the theoretical description of neutron - star envelopes with strong magnetic fields , but new challenges are put forward by observations of magnetars . in sect .
[ sect:1 ] we briefly overview recent work on the construction of models of neutron star atmospheres with strong magnetic fields and on the modeling of spectra of thermal radiation formed in an atmosphere or at a condensed surface .
we list important unsolved theoretical problems which arise in this modeling . in sect .
[ sect:2 ] , after a brief review of the effects of strong magnetic fields on the thermal structure and effective temperature of neutron stars , we describe our new calculations of the thermal structure .
compared to the previous results @xcite , we have taken into account neutrino energy losses in the outer crust of the star .
we show that neutrino emission strongly affects the temperature profile in a sufficiently hot neutron star and places an upper limit on its surface temperature @xmath5 and photon luminosity @xmath6 .
it was realized long ago @xcite that a neutron star atmosphere model should properly include the effects of a strong magnetic field and partial ionization . models of _ fully ionized _
neutron star atmospheres with strong magnetic fields were constructed by several research groups ( e.g. , * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* and references therein ) .
the most recent papers highlighted the effects that can be important for atmospheres of magnetars : the ion cyclotron feature @xcite and vacuum polarization , including a conversion of normal radiation modes propagating in the magnetized atmosphere @xcite .
early studies of _ partial ionization _ in the magnetized neutron star atmospheres ( e.g. , rajagopal , romani , & miller @xcite ; reviewed by @xcite ) were based on an oversimplified treatment of atomic physics and nonideal plasma effects in strong magnetic fields . at typical parameters ,
the effects of thermal motion of bound species are important .
so far these effects have been taken into account only for hydrogen plasmas .
thermodynamic functions , absorption coefficients , the dielectric tensor and polarization vectors of normal radiation modes in a strongly magnetized , partially ionized hydrogen plasma have been obtained and used to calculate radiative opacities and thermal radiation spectra ( see * ? ? ?
* and references therein ) .
the summary of the magnetic hydrogen atmosphere models and the list of references is given by @xcite .
the model is sufficiently reliable at @xmath7 g @xmath8 g , i.e. , in the field range typical of isolated radio pulsars .
it provides realistic spectra of thermal x - ray radiation @xcite .
@xcite extended this model to higher @xmath4 .
however , there remain the following unsolved theoretical problems that prevent to obtain reliable results beyond the indicated field range . *
the calculated spectra at @xmath9 g depend on the adopted model of mode conversion owing to the vacuum resonance and on the description of the propagation of photons with frequencies below the plasma frequency .
neither of these problems has been definitely solved .
their solution is also important for modeling the low - frequency ( uv and optical ) tail of the spectrum .
* at low @xmath10 or high @xmath4 , hydrogen atoms recombine in h@xmath11 molecules and eventually form a condensed phase ( see sect . [
sect : surf ] ) .
corresponding quantum - mechanical data are very incomplete .
* at @xmath12 g , transition rates of moving h atoms have not been calculated because of their complexity .
there is the only one calculation of the energy spectrum of bound states appropriate to this range of @xmath4 @xcite .
* a more rigorous treatment of radiative transfer in the atmosphere requires solving the transfer equations for the stokes parameters which has not been done so far for partially ionized atmospheres ( see , e.g. , @xcite for the cases of fully ionized atmospheres ) .
finally , we note that it is still not possible to calculate accurate atmospheric spectra at @xmath13 g for chemical elements other than hydrogen , because of the importance of the effects of motion of atomic nuclei in the strong magnetic fields .
apart from the h atom , these effects have been calculated only for the he atom @xcite , which _ rests _ as a whole , but has a moving nucleus , and for the he@xmath14 ion ( @xcite ; pavlov & bezchastnov @xcite ) .
the data of astrophysical relevance for he@xmath14 are partly published and partly in preparation ( see * ? ? ?
* ) ; one expects to have a he / he@xmath14 magnetic atmosphere model available in the near future .
the notion that an isolated magnetic neutron star has a condensed surface was first put forward by @xcite , who considered the iron surface . @xcite and @xcite studied the phase diagram of strongly magnetized hydrogen and showed that , when the surface temperature @xmath5 falls below some critical value ( dependent of @xmath4 ) , the atmosphere can undergo a phase transition into a condensed state .
a similar phase transition occurs for the equation of state of partially ionized , nonideal , strongly magnetized hydrogen plasma , constructed by potekhin , chabrier , & saumon ( @xcite ) for @xmath15 g and extended by @xcite to the magnetar field strengths .
it is analogous to the `` plasma phase transition '' suggested in plasma physics at @xmath16 ( see , e.g. , chabrier , saumon , & potekhin @xcite for discussion and references ) . according to potekhin et al .
( @xcite ) , the critical point for the phase transition in the hydrogen plasma is located at the density @xmath17 g @xmath3 and temperature @xmath18 k , where @xmath19 g. at @xmath20 the density @xmath21 of the condensed phase increases up to a few times of @xmath22 . on the other hand , according to @xcite , the surface density of a condensed phase for heavy elements is @xmath23 g @xmath3 , where @xmath24 and @xmath25 are the charge and mass numbers of the ions .
these two estimates of @xmath21 are in qualitative agreement .
@xcite estimates the critical temperature of hydrogen as @xmath26 k ( @xmath27 being the cohesive energy ) , also in agreement with the above estimate .
@xcite calculated the cohesive energy for ne and fe at @xmath7 g@xmath28 g , using the density functional theory ( dft ) and obtained @xmath29 kev .
recently , @xcite performed dft calculations of the cohesive energies for zero - pressure condensed hydrogen , helium , carbon , and iron at @xmath7 g @xmath30 g. for instance , they found that the cohesive energy per carbon atom ranges from @xmath31 ev at @xmath32 g to 20 kev at @xmath33 g. the cohesive energy per iron atom varies from @xmath34 kev at @xmath35 g to 33 kev at @xmath33 g. these calculations suggest @xmath36 of the same order of magnitude as the above estimate for hydrogen .
note that the models of @xcite and @xcite are constructed in the framework of the `` chemical picture '' of plasmas , whose validity near the plasma phase transition can be questionable @xcite .
thus the position ( and the very existence ) of the condensed surface requires further theoretical investigation and experimental or observational verification .
hopefully , this can be done by analyzing observations of thermal emission from neutron stars .
the thermal emission from the magnetized surface was studied by @xcite , turolla , zane , & drake ( @xcite ) , @xcite , and @xcite .
the spectrum exhibits modest deviations from blackbody across a wide energy range , and shows mild absorption features associated with the ion cyclotron frequency ( energy @xmath37 ev ) and the electron plasma frequency ( energy @xmath38 ev , where @xmath39 is in g @xmath3 ) .
however , the predictions of the ion cyclotron feature and the spectrum at lower frequencies are not firm .
the uncertainty arises from motion of the ions in the electromagnetic field around their equilibrium lattice positions .
most of the models treat the ions as fixed ( non - moving ) .
only @xcite considered two limits of fixed and free ions . in reality , however , the ions are neither fixed nor completely free ( see @xcite for estimates of possible uncertainties ) .
in addition , the condensed surface of a neutron star can be surrounded by a `` thin '' atmosphere , which is transparent to x - rays , but optically thick at lower wavelengths .
such a hypothesis has been first invoked by @xcite for explaining the spectrum of the isolated neutron star rx j0720.4@xmath403125 .
recently , the hydrogen atmosphere model , described in sect .
[ sect : atm ] , together with the condensed surface emission model of @xcite have been successfully used for fitting the spectrum of the isolated neutron star rx j1856.5@xmath403754 @xcite ( assuming the atmosphere to be `` thin '' as defined above ) .
the link between the magnetized atmosphere and stellar interior is provided by a _ heat blanketing ( insulating ) envelope_. the solution of heat transport problem relates the effective surface temperature @xmath5 to the temperature @xmath41 at the inner boundary of the blanketing envelope . without a magnetic field , it is conventional to place the inner boundary at @xmath42 g @xmath3 . in this case , it can be treated in the quasi - newtonian approximation with fractional errors @xmath43 @xcite . a strong magnetic field , however ,
greatly affects heat transport and , consequently , the thermal structure of the envelope .
the thermal structure of neutron star envelopes with radial magnetic fields ( normal to the surface ) was studied by @xcite ( also see @xcite for references to earlier work ) .
his principal conclusion was that the field reduces the thermal insulation of the heat blanketing envelope due to the landau quantization of electron motion .
the thermal structure of the envelope with magnetic fields normal and tangential to the surface was analyzed by @xcite and @xcite .
the tangential field increases the thermal insulation of the envelope , because the larmor rotation of electrons reduces the transverse electron thermal conductivity .
the case of arbitrary angle @xmath44 between the field lines and the normal to the surface was studied by @xcite in the approximation of constant ( density and temperature independent ) longitudinal and transverse thermal conductivities .
the authors proposed a simple formula which expresses @xmath5 at arbitrary @xmath44 through two values of @xmath5 calculated at @xmath45 and @xmath46 .
the case of arbitrary @xmath44 was studied also by @xcite and @xcite .
@xcite reconsidered the thermal structure of blanketing iron envelopes for any @xmath44 , using improved thermal conductivities @xcite .
@xcite analyzed accreted blanketing envelopes composed of light elements . in agreement with an earlier conjecture of @xcite and simplified treatments of @xcite and @xcite
, they demonstrated that the dipole magnetic field ( unlike the radial one ) does not necessarily increase the total stellar luminosity @xmath6 at a given @xmath41 . on the contrary ,
the field @xmath47@xmath48 g lowers @xmath6 , and only the fields @xmath9 g significantly increase it .
@xcite shifted the inner boundary of the blanketing envelope to the neutron drip , @xmath49 g @xmath3 , because in some cases they found a non - negligible temperature drop at @xmath50 g @xmath3 .
they obtained that magnetized accreted envelopes are generally more heat - transparent than non - accreted ones ( the same is true in the field - free case , studied by @xcite ) .
however , this heat transparency enhancement is less significant , when the transparency is already enhanced by a superstrong magnetic field .
recently @xcite showed that qualitatively the same dependence of @xmath6 on @xmath4 and on the chemical composition holds not only for dipole , but also for small - scale field configurations .
@xcite studied heat - blanketing envelopes with a magnetic field anchored either in the core or in the inner crust of the star ( with dipole and toroidal field components of different strengths ) .
they showed that a superstrong field in the inner crust of a not too hot star can significantly affect the surface temperature distribution and make it nonuniform and even asymmetric , with hot spots having different temperatures .
similar results were obtained by @xcite , who also evaluated pulsed fractions and phase - dependent spectra of neutron stars with strong magnetic fields anchored in the inner crust .
magnetars differ from ordinary pulsars in two respects : they possess superstrong surface magnetic fields , and they are generally younger and hotter .
the first circumstance suggests to extend the heat - blanketing layer to deeper layers ( at least to the neutron drip density as was done by @xcite ) .
the second indicates that neutrino emission can be important in the heat - blanketing envelopes .
accordingly , we have modified our computer code , which calculates the thermal structure @xcite , to make it fully relativistic and to allow for energy sinks during heat diffusion .
a complete set of equations for mechanical and thermal structure of a spherically symmetric star in hydrostatic equilibrium has been derived by @xcite .
they can easily be transformed to the form valid in an envelope of a star with radial heat transport , anisotropic ( slowly varying ) temperature distribution over any spherical layer , and a force - free magnetic field .
assuming quasistationary heat transport and neutrino emission , these equations reduce to the following system of ordinary differential equations for the metric function ( gravitational potential ) @xmath51 , the local heat flux @xmath52 , temperature @xmath10 , and gravitational mass @xmath53 , contained within a sphere of circumferential radius @xmath54 , as functions of pressure @xmath55 : @xmath56 here @xmath39 is the mass density ( equivalent energy density of the matter ) , @xmath57 is the net energy loss per unit volume ( @xmath58 is the neutrino emissivity in our case , although generally @xmath59 , @xmath60 being the heat deposition rate , e.g. , due to nuclear reactions ) , @xmath61 , @xmath62 is the opacity , @xmath63 is the so - called `` baryon mass density , '' @xmath64 is the baryon number density , @xmath65 is the mass of the hydrogen atom , @xmath66 is the stefan - boltzmann constant , @xmath67 is the local gravity , and @xmath68 is the gravitational constant . furthermore , @xmath69 are general relativistic corrections to radius , enthalpy and surface gravity , respectively .
we adopt the conventional definition of the opacity @xmath62 , used also by @xcite . in our notations
@xmath70 , where @xmath71 is the total ( electron plus radiative ) thermal conductivity of the plasma .
note that @xmath72 and @xmath73 in the entire neutron star envelope .
we use the same thermal conductivities as in @xcite .
the effective radial thermal conductivity in a local part of the surface equals @xmath74 , where @xmath75 and @xmath76 are the components of the conductivity tensor responsible for heat transport along and across field lines , respectively .
the local ( non - neutrino ) luminosity equals the integral of the flux over the sphere of radius @xmath54 , @xmath77 where @xmath78 are the polar and azimuthal angles . for a magnetic dipole model @xcite , @xmath79
the boundary conditions to eqs .
( [ phi])([m ] ) are @xmath80 and the surface pressure is determined , following gudmundsson , pethick , & epstein ( @xcite ) , by the condition @xmath81 @xmath82 , where @xmath83 is the radiative opacity .
the subscript ` s ' refers to surface values ; @xmath84 is the gravitational stellar mass . for the cases shown in fig .
[ fig : t1 ] .
top panel : @xmath32 g , bottom panel : @xmath85 g ; solid lines
ground - state matter , dot - dashed lines
@xmath86fe , dashed lines without neutrino emission for ground - state matter.,scaledwidth=37.0% ] general relativity correction factors @xmath87 , @xmath88 , and @xmath89 in eqs .
( [ k_r])([k_g ] ) are nearly constant because @xmath90 and @xmath91 at the bottom of the outer crust .
however , they are taken into account in our code , in order to extend calculations to deeper neutron star layers , when required .
equations ( [ l_r ] ) , ( [ t ] ) are one - dimensional , which implies that the mean temperature gradient along stellar radius is large compared to the tangential temperature gradient , i.e. , @xmath92 , where @xmath93 is a coordinate along the stellar surface .
let us roughly estimate the mean value of @xmath94 for a large - scale ( e.g. , dipole ) magnetic field , following @xcite .
in this case , @xmath95 , where @xmath96 is the depth at which the temperature distribution becomes nearly isotropic ( @xmath97 at the depth @xmath96 ) and @xmath98 is the mean surface temperature .
at the bottom of the outer crust ( i.e. , assuming @xmath99 ) we have @xmath100 ( e.g. , @xmath101 km for @xmath102 and @xmath103 km ) and @xmath104 ( see sect .
[ sect : res ] ) .
this estimate gives @xmath105 .
nevertheless , the one - dimensional approximation is inaccurate at those loci where magnetic field lines are tangential to the surface , because in this case one can not neglect their curvature , @xmath106 .
for large - scale magnetic field ( @xmath107 ) , the maximum size @xmath108 of such sites can be estimated as the distance at which an initially tangential field line crosses the depth @xmath96 ( which assumes that heat flows along field lines , i.e. , @xmath109 ) . thus ,
since @xmath111 is minimal at such sites , their contribution to the total stellar luminosity can be neglected in the first approximation . from plasmon decay ( solid lines ) , electron neutrino bremsstrahlung ( dot - dashed lines ) , and electron synchrotron radiation ( dashed lines ) . left panels : @xmath32 g ( radial field )
; right panels : @xmath85 g ; top panels : @xmath112 k ; bottom panels : @xmath113 k. , scaledwidth=48.0% ] we have solved eqs . ( [ phi])([m ] ) using a straightforward generalization of the runge - kutta method employed in our previous papers ( @xcite ; potekhin & yakovlev @xcite ; @xcite ) .
temperature profiles have been calculated within a local part of the blanketing envelope with a locally constant magnetic field . at every value of @xmath55 , a corresponding value of @xmath39 was found from the equation of state of magnetized electron - ion relativistic plasma ( e.g. , @xcite ) , using approximations of fermi - dirac integrals presented in @xcite . in all examples shown in the figures we chose the `` canonical '' neutron star mass @xmath102 and radius @xmath103 km .
the neutrino emissivity is calculated as @xmath114 , where the contributions due to electron - positron pair annihilation @xmath115 , plasmon decay @xmath116 , synchrotron radiation of neutrino pairs by electrons @xmath117 , and bremsstrahlung in electron - nucleus collisions @xmath118 are given , respectively , by eqs .
( 22 ) , ( 38 ) , ( 56 ) , and ( 76 ) of @xcite .
according to the results of @xcite and @xcite , other neutrino emission mechanisms are unimportant in the outer crust of neutron stars .
our calculations show that neutrino emission is crucially important for the thermal structure of neutron stars with internal temperature @xmath119 k. figure [ fig : t1 ] shows temperature profiles at @xmath120 k , @xmath121 k , and @xmath122 k for magnetic fields
@xmath32 g and @xmath33 g , directed perpendicular to the stellar surface .
the present results ( solid and dot - dashed lines ) are compared with the profiles calculated neglecting neutrino emission ( dashed lines ) . at the lowest temperature @xmath120 k
there is virtually no difference ( all the lines coincide ) . at @xmath123
k , the difference is noticeable , and at @xmath124 k it is large .
because of the growth of neutrino emission with increasing temperature at @xmath2 g @xmath3 , @xmath5 is nearly independent of @xmath41 at @xmath125 k , but depends on the magnetic field . in this figure
we also compare temperature profiles for the different heavy - element compositions of the outer envelope : iron ( dot - dashed lines ) and ground - state matter @xcite .
the effect of composition is not strong , but noticeable when the neutrino emission is important , because @xmath126 depends on the electron number density that is a function of composition for a given @xmath39 .
figure [ fig : f ] demonstrates the profile of the local heat flux @xmath52 , for the same cases as in fig .
[ fig : t1 ] , plotted by the same line types . without neutrino emission ( dashed lines ) , @xmath52 would be nearly constant , with only @xmath127% increase towards the inner crust due to the general relativity effects ( associated with the variation of the metric function @xmath51 ) and @xmath128% increase because of the spherical geometry ( the @xmath129 factor ) .
the neutrino emission leads to a strong dependence of the flux on @xmath39 and violates the familiar relation between @xmath52 and @xmath5 derived in the absence of energy sinks .
figure [ fig : q ] shows which neutrino emission mechanism dominates at given @xmath39 , @xmath41 , and @xmath4 . for superstrong magnetic fields , the neutrino synchrotron mechanism dominates in certain density ranges , which does not happen at `` ordinary pulsar ''
@xmath130 g. it is natural because of the strong @xmath4-dependence of the synchrotron neutrino emissivity .
pair annihilation neutrino emissivity is not seen in the figure , because it is too small .
notice that the emissivity of plasmon decay and bremsstrahlung processes can be affected by superstrong magnetic fields which has not been explored so far ( and we present the emissivities in the field - free case ) .
a slow dependence of these emissivities on @xmath4 , seen in fig . [ fig : q ] , is indirect ( caused by the dependence of temperature profiles on @xmath4 ) .
figures [ fig : pfs12 ] and [ fig : pfs15 ] give the @xmath131 relation for the magnetic fields @xmath32 g and @xmath85 g perpendicular and parallel to the radial direction .
the relations obtained with and without allowance for neutrino emission are plotted by solid and dashed lines , respectively .
we see that at @xmath132 k the neutrino emission does not affect @xmath5 . at higher @xmath119
k , in contrast , this emission is crucial : if @xmath133 , then @xmath5 continues to grow up with increasing @xmath41 , whereas with realistic @xmath126 the surface temperature tends to a constant limit , which depends on @xmath44 and @xmath4 . in most cases
this limit is reached when @xmath134 k , but for a superstrong field ( right panel ) and transverse heat propagation , it is reached at still smaller @xmath135 k. as a function of temperature @xmath41 at the neutron drip point for a neutron star with the magnetic field @xmath32 g , directed along stellar radius ( @xmath45 , parallel heat transport , sign @xmath136 ) or along the surface ( @xmath137 , transverse transport , @xmath138 ) .
ground - state composition is assumed .
solid lines
present calculation , dashed lines
neutrino emission is neglected.,scaledwidth=40.0% ] in fig .
[ fig : t0 ] we explore joint effects of magnetic field , neutrino emission , and the shift of the inner boundary from @xmath139 g @xmath3 to the neutron drip . here
magnetic field lines are directed along the surface ( perpendicular to the direction of heat transport ) .
therefore , these effects are most pronounced .
the solid lines are the results of accurate calculations ; the dashed lines , as before , show the model with @xmath133 ; and dot - dashed lines are obtained by solving eqs .
( [ phi])([m ] ) in the whole domain @xmath140 g @xmath3 , but with the magnetic field artificially `` switched off '' at @xmath50 g @xmath3 .
this is a simulation of the model , where the heat transport in the magnetized plasma is solved accurately up to @xmath141 g @xmath3 , while after this boundary nonmagnetic heat balance and transport equations are solved . in the absence of the results reported here , the latter model was used by @xcite to study the thermal structure and evolution of magnetars as cooling neutron stars with a phenomenological heat source in a spherical internal layer . in the left panel of fig .
[ fig : t0 ] , the solid and dot - dashed lines almost coincide , indicating that in this case @xmath141 g @xmath3 may provide a sufficient accuracy . in the right panel ,
in contrast , the effect of the shift of the inner boundary is quite visible .
therefore , we conclude that the development of our thermal structure code , reported here , will allow us to study the thermal history of magnetars at a higher accuracy level . meanwhile , a comparison of the profiles shown in figs .
[ fig : t1 ] and [ fig : t0 ] prompts that at @xmath142 k and @xmath143 g the magnetic effects on the conductivity could be important at still higher densities in the inner crust .
investigation of this possibility requires taking into account the effects of free ( possibly superfluid ) neutrons on thermal conduction and neutrino emission .
we are planning to perform such study in the future . , but for @xmath85 g.,scaledwidth=40.0% ] in the above figures we have shown the results of calculations where the surface of a neutron star was assumed diffuse , i.e. , without the phase transition discussed in sect .
[ sect : surf ] .
however , for @xmath85 g and @xmath144 k the surface temperature @xmath5 is below the estimates of the critical temperature @xmath36 given in sect .
[ sect : surf ] .
this possibility is explored in fig .
[ fig : pfc ] . here
we use the equation of state containing nonideal terms for strongly magnetized fully ionized plasma ( sect .
iiib of @xcite ) , which enforce the phase transition . at @xmath85
g the surface density is very high , @xmath145 g @xmath3 , in our model ( and the analytic estimate in sect .
[ sect : surf ] gives @xmath146 g @xmath3 ) .
the solid lines in the left panels reproduce the profiles shown in figs . [
fig : t1 ] and [ fig : t0 ] , whereas the dot - dashed lines display the case of magnetic condensation . in the right panels ,
we show the same temperature profiles as a function of local proper depth @xmath147 ( @xmath148 ) , measured from the radiative surface .
although the temperature profiles with and without magnetic condensation are drastically different in the surface layers , the effective temperature remains almost the same .
thus , the magnetic condensation does not significantly affect the @xmath131 relation .
this should not be surprising , because , as explained by @xcite , the main regulator of the @xmath41@xmath5 relation is the `` sensitivity strip '' where @xmath71 has a minimum .
this domain , a real bottleneck for the heat outflow , lies at @xmath149 ( except for low @xmath5 , not considered here ) , and therefore it is not affected by the condensation .
we have described the main effects of strong magnetic fields on the properties of neutron star atmospheres and heat blanketing envelopes , with the emphasis on the difference between ordinary neutron stars and magnetars .
observations of magnetars pose new theoretical problems and challenges because of magnetars superstrong magnetic fields and high temperatures .
we also report a solution to one of such problems , which consists in taking into account neutrino energy losses in the outer crust of hot , strongly magnetized neutron stars . we have demonstrated that , because of these losses , the effective surface temperature @xmath5 almost ceases to depend on the temperature @xmath41 in the inner crust as soon as @xmath41 exceeds @xmath121
k. a direct consequence of this observation is that in the absence of powerful energy sources in outer envelopes , the stationary ( time - averaged ) effective temperature can not be raised above the value that it would have at @xmath150 k , irrespectively of the energy release in the deeper layers .
we are grateful to the anonymous referee for useful comments .
we thank a.d.kaminker for pointing out to some misprints in formulae , which are corrected in this electronic version of the paper .
ayp thanks the organizers of the conference `` isolated neutron stars : from the interior to the surface '' ( london , april 2428 , 2006 ) , especially silvia zane and roberto turolla , for perfect organization , attention and support .
fruitful discussions with participants of this conference have significantly contributed to the developments partly reported in the present paper .
g is directed parallel ( top panels ) or perpendicular ( bottom panels ) to the stellar surface ;
@xmath123 k. solid lines the model of diffuse surface ( no phase transition ) , dot - dashed lines condensed surface.,scaledwidth=48.0% ] pavlov g.g . , shibanov yu.a . ,
zavlin v.e . , &
meyer r.d . : neutron star atmospheres . in : alpar m.a .
, kizilolu . , van paradijs , j. ( eds . ) the lives of the neutron stars , proceedings of the nato asi ser .
c , * 450 * , p. 71 .
kluwer , dordrecht ( 1995 ) zavlin v.e .
& pavlov g.g . : modeling neutron star atmospheres . in : becker w. , lesch h. , trmper j. ( eds . ) proceedings of the 270 .
we - heraeus seminar on neutron stars , pulsars , and supernova remnants , mpe report 278 , p. 263 .
mpe , garching ( 2002 ) | strong ( @xmath0 g ) and superstrong ( @xmath1 g ) magnetic fields profoundly affect many thermodynamic and kinetic characteristics of dense plasmas in neutron star envelopes .
in particular , they produce strongly anisotropic thermal conductivity in the neutron star crust and modify the equation of state and radiative opacities in the atmosphere , which are major ingredients of the cooling theory and spectral atmosphere models . as a result
, both the radiation spectrum and the thermal luminosity of a neutron star can be affected by the magnetic field .
we briefly review these effects and demonstrate the influence of magnetic field strength on the thermal structure of an isolated neutron star , putting emphasis on the differences brought about by the superstrong fields and high temperatures of magnetars .
for the latter objects , it is important to take proper account of a combined effect of the magnetic field on thermal conduction and neutrino emission at densities @xmath2 g @xmath3 .
we show that the neutrino emission puts a @xmath4-dependent upper limit on the effective surface temperature of a cooling neutron star . |
the soft - collinear effective theory ( scet ) constructed in @xcite offers a systematic description of processes involving energetic particles .
it has an expansion in a small parameter @xmath10 , where @xmath11 is a typical transverse momenta and @xmath12 the large energy scale .
hard exclusive and inclusive processes in qcd are usually described using the powerful techniques of qcd factorization and light - cone expansions @xcite .
scet encompasses and extends these frameworks , and in particular allows a model independent description of effects caused by the interplay between energetic collinear particles and soft particles beyond leading order in the power expansion .
these effects can be described in a rigorous way based solely on qcd , but are not included in purely collinear expansions .
the study of operators that describe these mixed collinear - ultrasoft ( collinear - usoft ) effects is the purpose of this paper . for recent applications of scet in hard scattering processes and b - decays
see refs .
@xcite . since our focus is on mixed collinear - usoft interactions , we consider collinear quark fields @xmath13 , collinear gluon fields @xmath14 , usoft heavy quark fields @xmath15 , usoft light quark fields @xmath16 , and usoft gluons @xmath17 .
( we follow the notation in refs .
@xcite , but for simplicity will often suppress the momentum label @xmath18 on the collinear fields . ) these degrees of freedom can interact in a local manner in lagrangians and currents .
this is in contrast with collinear - soft couplings whose interactions are mediated by offshell fluctuations @xcite , and appear in external operators .
we comment on collinear - soft interactions at the end of the paper .
the derivation of the leading order collinear quark and gluon lagrangians @xmath19 and @xmath20 can be found in ref .
@xcite , and a description of the gauge symmetries of scet can be found in refs .
@xcite . for details on power
counting we refer to ref .
the heavy - to - light currents at large energy , @xmath21 , were derived to leading order in ref .
@xcite , including one - loop matching for all the wilson coefficients .
the running of these wilson coefficients was considered in refs .
@xcite . in the context of the scet , power
suppressed corrections were first considered in ref .
@xcite , and the @xmath1 suppressed currents @xmath21 and collinear quark lagrangians were derived .
the authors showed that a reparameterization invariance ( rpi ) uniquely fixes the wilson coefficients of their subleading currents and lagrangian in terms of the leading order coefficients .
was used to derive constraints on matrix elements @xmath22 . ] in ref .
@xcite the rpi of scet was extended to the most general three classes ( i , ii , iii ) , and the multipole expansion of the collinear quark lagrangian was treated to higher orders in @xmath23 and were shown not to receive anomalous dimensions . in ref .
@xcite the presence of additional @xmath1 heavy - to - light currents was pointed out that were missing in ref .
@xcite .
the study of power corrections in scet was continued in ref .
@xcite and several important results were obtained for mixed usoft - collinear operators .
in particular the mixed usoft - collinear quark lagrangian @xmath24 was first considered and was derived to @xmath25 working at tree level , but to all orders in attachments of @xmath26 gluon fields . in a similar fashion heavy - to - light currents were derived to @xmath0 , and linear combinations of currents that are invariant under the three types of rpi were identified .
it was also shown that the operators in @xmath24 are not renormalized based on an analysis of arbitrary @xmath27-loop diagrams in the hard region of qcd .
the mixed usoft - collinear quark lagrangian @xmath24 was extended to a gauge invariant form with covariant derivatives in ref .
@xcite .
the purpose of the present paper is to answer some open questions regarding our knowledge of the power suppressed usoft - collinear lagrangian and heavy - to - light currents .
this includes the number of @xmath21 currents at @xmath1 , since even at tree level the full reduction of dirac structures has not yet been implemented . for both @xmath21 and @xmath24 we also construct a complete basis which is valid for matching at any order in @xmath2 , and therefore includes all operators that can be induced by radiative corrections or operator mixing .
we work in the most general possible frame throughout ( eg . allowing @xmath28 ,
@xmath29 ) , and consider all the restrictions from rpi including the transformation of wilson coefficients .
finally , we include the mixed usoft - collinear pure glue lagrangian beyond lo ( which follows from an extension of work in refs .
the above results are obtained by considering the full implications of rpi , and including all possible operators allowed from collinear gauge invariance , power counting , and the reduction of dirac structures from the effective theory fields .
we restrict ourselves to @xmath1 which is one order less than the order to which the tree - level matching results are known from ref .
we treat @xmath24 to @xmath0 , and give a detailed account of how the gauge invariant form in ref .
@xcite was derived . in cases
where our results are restricted to those in refs .
@xcite we find agreement , as discussed in more detail in the body of the paper .
the results derived here are sufficient for the proof of a factorization theorem for heavy - to - light form factors to all orders in @xmath2 and leading order in @xmath30 @xcite . ] for the heavy - to - light currents at @xmath1 an important result we find is a new type of `` three - body '' currents , which have not been previously considered in the literature . in refs .
@xcite the attention was restricted to scet operators of two - body type @xmath31 , where the two products in parenthesis are collinear gauge invariant , and the ellipses denote combinations of collinear derivatives . beyond tree level but at the same order in @xmath23 , we find that three - body structures can appear for some of the currents , having the form @xmath32 with three collinear gauge invariant factors .
we show the rpi can be used to determine for which currents this happens .
we also show that rpi greatly restricts the form of the three - body operators , so that they always involve a collinear gluon field strength .
the two - body operators have hard wilson coefficients which are functions of a single parameter @xmath33 , while the new three - body operators have two parameter coefficients @xmath34 .
analogous three - body structures could appear in the usoft - collinear lagrangian @xmath24 at higher orders in perturbation theory , however using constraints from symmetries of scet we prove that this does not occur .
our results are relevant to the study of decay channels for @xmath35 mesons which involve energetic hadrons in the final state .
for instance , the results derived in this paper are necessary ingredients in the factorization formula for heavy - to - light form factors proven in ref .
@xcite ( for earlier work on factorization in heavy - to - light form factors see ref .
@xcite , and for results from qcd sum rules see refs .
the factorization theorem is valid to all orders in @xmath2 and leading order in @xmath30 , @xmath36 , and separates contributions from the scales @xmath37 , @xmath38 , and @xmath39 , where @xmath40 is a hadronic scale .
it states that a generic form factor can be split into two types of contributions @xmath41 where @xcite @xmath42 @xmath43 , and the two terms both scale as @xmath44 .
this scaling is model independent and is in agreement with that derived from qcd sum rules @xcite . in eq .
( [ ffintro ] ) @xmath45 and @xmath46 are standard nonperturbative light - cone distribution amplitudes , c.f
. @xcite .
the hard coefficients @xmath47 and @xmath48 can be calculated in an expansion in @xmath49 and are simply related to the wilson coefficients of the @xmath50 current operators @xmath21 .
the jet function @xmath51 is dominated by momenta @xmath52 .
if we wish to expand in @xmath53 then using the techniques developed in ref .
@xcite , @xmath51 is _ calculable _ in terms of time - ordered products of the @xmath54 operators @xmath21 and @xmath24 that we study here . at tree - level [ ie .
@xmath55 one finds that @xmath51 contains a @xmath56 , and in ratios of form factors the results for @xmath57 then agree with terms computed in ref .
the @xmath58 dependence first shows up at @xmath59 as does possible dependence on @xmath60 .
however , as we show in section [ bpisection ] it is possible to absorb the @xmath60 terms into a redefinition of the @xmath61 to all orders in perturbation theory .
the factorization formula provides a clean separation of the `` soft '' non - factorizable ( nf ) contributions and `` hard '' factorizable ( f ) terms without double counting .
it also gives us a procedure to systematically improve the predictions to any order in perturbation theory at leading order in @xmath30 .
the value of @xmath48 and @xmath47 can depend on which heavy - to - light process we consider , whereas @xmath45 and @xmath62 are universal functions .
the @xmath63 s are also universal since only a @xmath64 appears for decays to pseudoscalars @xmath65 , and a @xmath66 and @xmath67 appear if @xmath65 is a vector meson .
the jet functions @xmath51 are common among certain classes of form factors and also do not depend on the precise state ( eg . @xmath68 or @xmath69 ) .
the @xmath70 terms satisfy the so - called large energy form factor relations @xcite , as expected from the prior loose definitions of these terms as `` soft '' contributions @xcite .
note that we have not bothered to separate @xmath71 and @xmath72 fluctuations in the @xmath61 functions , since it is not clearly beneficial phenomenologically .
the factorization theorem does tell us that @xmath73 , however it does not distinguish between factors of @xmath74 and @xmath75 in this @xmath76 .
it also does not numerically favor the @xmath57 or @xmath70 term , for instance it is possible that the leading @xmath53 in @xmath51 is compensated by an analogous factor in @xmath61 .
we start in section ii by reviewing the general constraints imposed on scet operators following from collinear gauge invariance , spin structure reduction , and reparameterization invariance . in section iii
we study the implications of these predictions for the subleading usoft - collinear lagrangian @xmath77 . in section
iv we present detailed results for scet currents .
using the example of the scalar current as the pedagogical example , we demonstrate the construction of the complete basis of @xmath78 operators contributing to the weak currents , which closes under rpi transformations .
explicit results are then also derived for the pseudo - scalar , vector , axial - vector , and tensor heavy - to - light currents to @xmath1 . in section
v we summarize the one - loop matching results for the currents , give explicit results for @xmath24 feynman rules , and discuss the basis of currents in the particular frame @xmath7 , @xmath79 . | we construct operators that describe power corrections in mixed collinear - ultrasoft processes in qcd .
we treat the ultrasoft - collinear lagrangian to @xmath0 , and heavy - to - light currents involving collinear quarks to @xmath1 including new three body currents .
a complete gauge invariant basis is derived which has a full reduction in dirac structures and is valid for matching at any order in @xmath2 .
the full set of reparameterization invariance ( rpi ) constraints are included , and are found to restrict the number of parameters appearing in wilson coefficients and rule out some classes of operators . the qcd ultrasoft - collinear lagrangian has two @xmath3 operators in its gauge invariant form . for the @xmath4 heavy - to - light currents there are @xmath5 subleading ( scalar , pseudo - scalar , vector , axial - vector , tensor ) currents , where @xmath6 have coefficients that are not determined by rpi .
in a frame where @xmath7 and @xmath8 the total number of currents reduces to @xmath9 , but the number of undetermined coefficients is the same .
the role of these operators and universality of jet functions in the factorization theorem for heavy - to - light form factors is discussed . |
the basic evolution of low mass x
ray binaries above the ` bifurcation period ' @xcite has been often described ( e.g. * ? ? ?
the endpoint of this evolution is a wide system ( orbital period from a few to hundreds of days ) containing a millisecond pulsar ( msp , the neutron star primary of initial mass m@xmath5 spun up by mass transfer from the secondary , of initial mass m@xmath6 ) and a low mass helium white dwarf , remnant helium core of the mass losing red giant .
the simple relation connecting the giant stellar radius with the helium core mass of the star , which has a very slight dependence on the stellar mass and on the chemical composition , coupled with the kepler law , produces a relation between the final white dwarf mass and the final , which has been often studied ( e.g. * ? ? ? * ; * ? ? ? * ; * ? ? ? * ) and compared with the vs. distribution of the more than 60 known msps .
the sample of binary msp known up today shows a shortage of systems with orbital periods between 22 and 56 days .
there have been two main attempts to explain this ` period gap ' . according to @xcite , the gap is related to the bifurcation period , which he assumes to be around 2 days .
systems of secondary mass m@xmath7m@xmath8 , and beginning mass transfer just above 2 days period , should evolve to final @xmath960 days .
if the secondary has an initial mass @xmath91.4 , the binary might evolve through a second common envelope phase , ending at p@xmath1020 days .
@xcite revise this picture attributing the periods shorter than 20 days to early massive case b evolution .
they question the idea that the bifurcation period is as large as 2 days ( to allow a minimum final period larger than 60 days ) , as evolution towards shorter periods for systems below ( but close to ) 2 days requires a very strong magnetic braking .
@xcite then attribute the period gap to the small range of initial core masses involved ( 0.17 @xmath11 ) . in this latter scenario , a carbon oxygen white dwarf of mass larger than 0.35should be present at periods shorter than 20 days .
notice that the two best determined masses are @xmath12 , at = 12.3d ( kaspi et al .
1994 ) and @xmath13 , at = 5.741d @xcite , so the standard case b evolution produces at least some systems below 20d .
we reconsider the evolution of the systems containing a neutron star and a giant of low initial mass .
we only consider the systems in which the initial secondary mass is smaller than the primary mass , in the range @xmath14 , and assume population i composition ( metal mass fraction z=0.02 ) . for a wide range of initial periods ,
the systems suffer a phase of detachment , when the hydrogen burning shell reaches the point of deepest penetration of the convective envelope at the basis of the red giant branch @xcite . when the hydrogen content suddenly increases , there is a thermal readjustment of the shell physical conditions , with a decrease in the shell temperatures and , consequently , in the nuclear burning luminosity .
thus the star contracts in order to adjust to the level of nuclear reactions in the shell .
this phase is an important evolutionary tool in the single star evolution , as the radius contraction is associated to a drop in the luminosity along the red giant branch , producing a bump " in the luminosity function of simple stellar populations , which has been observed in many globular clusters @xcite .
it has also been often identified in binary evolution @xcite , although never discussed in the context of the evolution leading to msp .
we show that the bump related detachment of the system occurs for a good fraction of the evolutionary paths leading to long period binary msp , and consider the possible consequences for the orbital period evolution .
although we do not follow explicitly the spin up evolution of the accreting neutron star , we consider that , when the system resumes mass transfer , the ns is now a millisecond pulsar .
therefore , mass accretion on the ns may be inhibited by the radio ejection due to the pressure exerted by the msp radiation on the matter at the inner lagrangian point @xcite .
this alters the binary evolution , as the mass is lost from the system carrying away a specific angular momentum ( am ) larger than the average .
in fact , the am lost should be of the order of magnitude of the am of the donor , whose mass at this stage is only 0.40.5 , much smaller than the ns mass .
we model parametrically the evolution in the phase of radioejection , and show that it naturally leads , in many cases , the final periods of systems away from the range 2060 days .
we also show that , if the onset of radio ejection is the mechanism which produces the period gap , this gap should not be found in population ii .
unfortunately , the orbital period distribution in globular cluster seems to be altered by destruction mechanisms at long orbital periods , so that this hypothesis by now can not be tested . as a byproduct
, we also show that the relation vs. obtained for population i in the framework of the computed models is consistent with the three best observational data .
we build up our stellar models by the aton2.1 code , whose input physics is described in @xcite , while the binary evolution routines follow the description in @xcite .
the code has been recently updated also including explicitly the evolution of the ns @xcite .
models of 1.1 , 1.2 and 1.3 , with helium mass fraction y=0.28 and metallicity z=0.02 , were evolved as donors in binaries containing a ns companion of initial mass in the range 1.21.35 .
the initial separation was chosen to allow mass transfer at different stages along the evolution .
we do not discuss the previous evolution of these putative systems , that is the formation of the neutron star via the supernova ( sn ) ejection , and whether it is possible to obtain a neutron star in circular orbit with a low mass companion at the periods from which we begin considering the mass transfer .
in particular , the common envelope phase preceding the sn event is likely to end with short orbital periods for low mass companions , and long period systems ( e.g. p@xmath15100d ) would be created mainly as a consequence of strong natal kicks ( see , e.g * ? ? ? * ) and have strong eccentricities .
in addition to nuclear evolution , we assume that magnetic braking is active and describe it according to @xcite , with the _ f_parameter fixed at 0.5 .
we also computed several evolutions of systems of m=0.9 and 1.1 , for y=0.24 and z=10@xmath16 to test the different behavior of population ii systems .
table 1 reports some parameters for the binary evolutions starting from different initial models along the evolution of the 1.1 as single star , labeled by the model number n@xmath17 .
smaller n@xmath17 thus correspond to earlier evolutionary stages .
we computed standard " evolutions , which , at least during the first phase of mass transfer , neglect the possible role of the pulsar .
a series of models ( labeled by con " in table 1 ) assume conservative mass transfer below the eddington limit .
the matter exceeding the eddington limit was assumed to be lost from the system , carrying away the specific am ( j@xmath18 ) of the ns .
we considered also evolutions , ( labeled by
eta " ) in which only a fraction @xmath19 of the mass lost by the donor was transferred to the ns , and the remaining half was assumed to leave the system , carrying away the specific am j@xmath18 .
we add several evolutions in which we start again from the bump related detachment , and assume that all the mass lost from the donor is lost from the system , taking away a large fraction ( from 60% to 100% ) of the specific am j@xmath6 of the donor : this is the way in which we simulate an efficient onset of radio ejection , as this mechanism stops the matter at the inner lagrangian point .
initial and final masses of the components and final periods are also reported in table 1 .
fig.[fig1 ] shows the final vs. relation for the population i evolutions of table 1 ( full squares on the left ) and for the additional population ii models ( full squares on the right ) , and compares them with the analytic approximations provided by @xcite and @xcite for population i and ii . our data compare well with the population ii relation by @xcite , but it is steeper than theirs at periods @xmath20days for population i. we notice that our population i results are in good agreement with the three observed data points displayed in the figure .
these same systems , however , have quite small neutron star masses : the central values for psr j0437 - 4715 @xcite and b1855 + 09 @xcite are only 1.551.60 , with upper limits at @xmath211.71.75 . in addition , the mass of psr j1713 + 0747 ( splaver et al . 2005 ) , at a period of 68d ( close to our evolutions
5 ) is only 1.3@xmath22 .
this value is three @xmath23 smaller than our predicted mass of @xmath24 from sequence 5con , while sequence 5eta provides a much more comfortable value of 1.655 , as we have assumed the loss of half of the mass lost from the donor , and , in addition , the initial mass of the ns is taken to be only 1.25 , contrary to the commonplace assumption of a starting value of 1.35 - 1.40 often adopted in the literature .
notice that the recent accurate mass determination of the relativistic double pulsar system psr j0737 - 3039 ( lyne et al .
2004 ) provides 1.25 for the lighter component . to maintain small masses for the pulsar component
, we could also have reduced the companion mass down to 1 .
when , on the contrary the initial mass of the donor is somewhat larger than 1.1 , the first phases will occur at mass transfer rates much larger than eddington limit , and the additional initial mass will easily be lost from the system . of course , there is no adequate physical reason to assume the semi conservative evolution we adopt , but a careful exploration of the parameter space is beyond the purpose of the present work and the results are not affected sensibly by the precise choices .
this is shown in fig [ fig2 ] , which compares the mass loss rate versus period evolution in cases 5con and 5eta . in this standard cases , after the super eddington inital phase of mass loss , a stationary phase begins , interrupted by the bump induced detachment , after which mass loss resumes and follows the same rules .
the difference between the conservative and non conservative evolution is not very important , due to the fact that we assume that the specific am loss associated with the mass loss is that of the primary neutron star , much smaller than the average am for most of the evolution .
the non conservative case , however , allows us to obtain in the end small neutron star masses ( 1.61.7 ) much more compatible with the neutron star masses in the systems we are examining .
the conservative evolution leads to masses 1.92.1 .
our following discussion depends on the orbital evolution and not on the assumption of mass conservation , as long as we assume am is lost from the neutron star .
fig . [ fig3 ] shows the hr diagram of the donor star 1.1for z=0.02 .
the evolutionary track without mass loss is shown as a reference .
the location of the bump is seen in the thickening of the track , at @xmath25 .
it corresponds to a stage at which the core mass is @xmath26 .
we see that a bump signature is present in all the evolutions where mass loss begins at evolutionary phases at which the hydrogen burning shell has not yet reached the discontinuity in hydrogen left by convection . at the bump related detachment , in all cases
, the ns has been accelerated by the first phase of mass transfer .
we assume that the maximum spin achieved is the observed minimum spin period @xmath27ms . during the semi
detached evolution , the total mass decreases , while the hydrogen shell advances and the h - exhausted core increases in mass , until it will encounter the discontinuity .
of course the earlier is the phase at which mass loss starts , the smaller is the total mass at the time of detachment .
when the mass loss rate is resumed , it is much larger if we assume that , thanks to an efficient radio ejection , the matter lost from the system carries away a large fraction of the specific am of the donor . figure [ fig4 ] shows the different evolutions obtained in cases 4re .
the larger is the fraction of specific am lost , the larger is the mass loss rate , and the shorter is the final .
we see that the standard evolution would lead to a final period of @xmath2136 days , in the middle of the period gap .
but if radio ejection is efficient the period is reduced below 22 days for a fraction between 60% and 80%@xmath28 of am loss , corresponding to a mass transfer rate of @xmath29/yr .
we must then understand when and where the mass transfer rates obtained are compatible with the assumption of radio ejection .
figure [ fig5 ] displays the mass transfer rates for cases 3 , 4 and 5 , and compares the standard evolution with the evolutions assuming that 80% of the donor am is lost starting after the bump related detachment .
contrary to the case discussed above , the evolution of case 5 leads to a final period in the middle of the gap ( @xmath2128d ) if there is an efficient radio - ejection , corresponding to a mass loss rate of 2@xmath30/yr , otherwise the final period is @xmath2161d , at the upper end of the period gap .
which are the constraints on the pulsar magnetic field which allow an efficient radio ejection in case 4 and non efficient in case 5 ?
we assume that the maximum spin frequency attainable by the msp is @xmath21750hz @xcite ( corresponding to a minimum period of @xmath211.3ms ) .
following burderi et al .
( 2001 ) , we remember that radio ejection can occur for orbital periods longer than the critical period : + + @xmath31^{-3/2 } \times \nonumber \\ ( m+m_2)^{-1/2 } \;\ ; { \rm h } & & \label{eq : pcrit } \end{aligned}\ ] ] where @xmath32 is the shakura sunyaev viscosity parameter , @xmath33 is the magnetic moment of the neutron star in units of 10@xmath34gauss @xmath3 ( corresponding to a surface magnetic field of @xmath35gauss ) , p@xmath36 is the spin period of the neutron star in milliseconds , @xmath37 for a gas with solar abundances ( where @xmath38 is the mean particle mass in units of the proton mass @xmath39 ) , @xmath40 is the accretion rate in units of @xmath41/yr .
this critical period has been derived equating the pressure of a shakura - sunyaev accretion disc ( approximately proportional to @xmath42 ) with the radiation pressure exerted by the magneto - dipole rotator ( proportional to @xmath43 ) .
this gives a radius beyond which radiation pressure always dominates over disk pressure , causing the ejection of matter of the disc .
equating this radius with the distance between the ns and the inner lagrangian point , which we approximate with the difference between the orbital separation and the roche lobe radius of the secondary ( which depends on the binary period ) , and solving for the orbital period gives a critical period beyond which accretion is always inhibited by the action of the pulsar pressure .
we can invert equation 1 to obtain the value of the critical magnetic momentum as a function of the critical orbital period : we obtain @xmath44 where @xmath45 , a value close to 1 , and @xmath46 . assuming that the nss have been accelerated to the minimum observed spin period @xmath47ms , inserting the appropriate values for the masses ( see table 1 ) , and assuming that the orbital period at the bump is the critical period
, we can derive a constraint on the magnetic moment of the ns .
namely , if we wish that the sequence 4 experiences radio ejection ( to reduce its final period below 22d ) while sequence 5 does not ( to avoid a final period into the gap ) , we must impose : @xmath48 for those systems whose initial binary parameters would lead them into the period gap . in this interpretation ,
the gap is only a statistical gap , and not a forbidden region .
therefore , we can predict that systems with into the gap can exist , but they should preferentially have magnetic moments out of the interval given in eq.[eq : two ] . fig .
[ figdata ] reports the orbital periods vs. the magnetic moments for all the known binary msps in the galactic field ( as derived from version 1.32 of the atnf pulsar catalog , and from stairs et al .
2005 ) having spin period shorter than 10 ms , @xmath49 d , nearly circular orbit ( eccentricity smaller than 10@xmath16 ) and supposedly white dwarf companion of mass @xmath50 m@xmath51 the aforementioned parameters are the typical outcomes of the standard nuclear evolution of low mass x - ray binary , which is in the focus of this paper ( e.g. * ? ? ? * ) .
it appears that about 75% of the sample displays magnetic moments within the range of equation [ eq : two ] or very close to its limits .
the evolution of the three systems at @xmath2113 days might have been influenced by radio ejection , and the two systems at @xmath2156 days could have been found in the gap , if their magnetic moments had been larger . the proposed scenario is finally summarized in fig .
[ fig6 ] , reporting the evolution in the plane vs. white dwarf mass .
we see that , taking into account the ns evolution and the process of radio ejection leads to a natural explanation of the period gap
. a population synthesis may give further support to this picture , allowing to precisely calculate the effect of the interplay between the various parameters in eq .
[ eq : pcrit ] on the vs @xmath52 distribution .
the result we have found derives from three main occurrences : 1 ) the maximum extension of the convective envelope in population i reaches down to an inner mass point of @xmath210.24 , so that the bump induced detachment occurs when the hydrogen exhausted core mass reaches about 0.22 , at a period well below 20 days for the systems in which we have tested that the radio ejection mechanism can work . in population ii ,
the maximum extent of convection only reaches @xmath210.35 .
thus any bump related detachment would occur at a much longer longer orbital period , when the core reaches @xmath53 and the red giant has already transfered a large amount of its envelope . therefore , even if radio ejection begins to dominate , it can not affect in an interesting way the binary final period . in any case
, no period gap can occur due to this mechanism , for periods shorter than @xmath2135 days .
unfortunately , the globular cluster period distribution @xcite does indicate that it is affected by the dynamic encounters in the dense stellar environment .
in particular , systems with period larger than 10 days are so few that we can not falsify our hypothesis by now .
we follow the evolution of the mass losing component in systems progenitors of binaries containing a msp and a remnant white dwarf companion .
when the mass transfer begins during the red giant branch phase , in many cases the binary evolution shows a period of detachment , due to the shrinking of the stellar radius which occurs when the hydrogen burning shell meets with the hydrogen discontinuity left at the time of the maximum extension of the convective envelope . when the mass transfer resumes , it is possible that the evolution suffers a radio ejection phase which alters the final orbital parameters , due to the loss of mass and am from the system .
we show that the period gap between 22 and 56 days in the distribution of binary msps is produced if the magnetic momentum of the neutron star is in the range @xmath54g@xmath3 , typical of msps .
a population synthesis study could strengthen the conclusion that this period gap is `` bump related '' , but there are many parameters to be considered in the analysis
. a better test of the model would be if new systems are found into the period gap , and their magnetic moments are either below @xmath55g@xmath3 , or above @xmath4g@xmath3 .
stairs , i. h. , faulkner , a. j. , lyne , a. g. , kramer , m. , lorimer , d. r. , mclaughlin , m. a. , manchester , r. n. , hobbs , g. b. , camilo , f. , possenti , a. , burgay , m. , damico , n. , freire , p. c. , gregory , p. c. 2005 , apj , in press , astro - ph/0506188 lcccccccccccl 2con & 1 & j@xmath18 & 200 & 0.0511 & 1.1 & 1.35 & & 0.194 & 2.256 & & 2.101 & -10.00 + 3con & 1 & j@xmath18 & 300 & 0.1550 & 1.1 & 1.35 & 0.293 & 0.246 & 2.072 & 8.52 & 13.20 & -9.27 + 4con & 1 & j@xmath18 & 400 & 0.1868 & 1.1 & 1.35 & 0.408 & 0.273 & 2.047 & 13.60 & 35.86 & -8.67 + 5con & 1 & j@xmath18 & 500 & 0.2049 & 1.1 & 1.35 & 0.505 & 0.290 & 2.012 & 16.80 & 61.63 & -8.38 + 6con & 1 & j@xmath18 & 600 & 0.2184 & 1.1 & 1.35 & 0.617 & 0.301 & 1.977 & 17.67 & 86.57 & -8.11 + 8con & 1 & j@xmath18 & 800 & 0.2383 & 1.1 & 1.35 & & 0.303 & 1.959 & & 105.0 & -8.05 + 9con & 1 & j@xmath18 & 900 & 0.2463 & 1.1 & 1.35 & & 0.310 & 1.939 & & 126.3 & -7.85 + 3eta & .5 & j@xmath18 & 300 & 0.1552 & 1.1 & 1.20 & & 0.244 & 1.628 & 8.14 & 12.10 & -9.27 + 4eta & .5 & j@xmath18 & 400 & 0.1868 & 1.1 & 1.35 & 0.415 & & 1.692 & 13.64 & & -8.80 + 5eta & .5 & j@xmath18 & 500 & 0.2049 & 1.1 & 1.25 & & 0.290 & 1.655 & 16.84 & 61.49 & -8.38 + 3re & 0 & 0.8j@xmath6 & 300 & 0.2106 & 0.293 & 2.025 & & 0.240 & 2.025 & & 9.817 & -8.84 + 4re06 & 0 & 0.6j@xmath6 & 400 & 0.2266 & 0.408 & 1.692 & & 0.263 & 1.692 & & 24.38 & -8.42 + 4re08 & 0 & 0.8j@xmath6 & 400 & 0.2266 & 0.408 & 1.912 & & 0.257 & 1.912 & & 19.35 & -8.08 + 4re1 & 0 & j@xmath6 & 400 & 0.2266 & 0.408 & 1.692 & & 0.253 & 1.692 & & 15.75 & -7.56 + 5re08 & 0 & 0.8j@xmath6 & 500 & 0.2324 & 0.505 & 1.796 & & 0.268 & 1.796 & & 28.28 & -7.77 + 5re1 & 0 & j@xmath6 & 500 & 0.2367 & 0.505 & 1.547 & & 0.261 & 1.547 & & 21.46 & -7.00 + | the monotonic increase of the radius of low mass stars during their ascent on the red giant branch halts when they suffer a temporary contraction .
this occurs when the hydrogen burning shell reaches the discontinuity in hydrogen content left from the maximum increase in the convective extension , at the time of the first dredge up , and produces a well known bump " in the luminosity function of the red giants of globular clusters .
if the giant is the mass losing component in a binary in which mass transfer occurs on the nuclear evolution time scale , this event produces a temporary stop in the mass transfer , which we will name bump related " detachment .
if the accreting companion is a neutron star , in which the previous mass transfer has spun up the pulsar down to millisecond periods , the subsequent mass transfer phase may be altered by the presence of the energetic pulsar .
in fact , the onset of a radio ejection phase produces loss of mass and angular momentum from the sytem .
we show that this sequence of events may be at the basis of the shortage of systems with periods between @xmath0 and @xmath1 days in the distribution of binaries containing millisecond pulsars .
we predict that systems which can be discovered at periods into the gap should have preferentially either magnetic moments smaller than @xmath2g@xmath3 , or larger than @xmath4g@xmath3 .
we further show that this period gap should not be present in population ii . |
the cobe experiment ( bennet et al . 1992 , 1994 ; smoot et al . 1992 ) has stimulated a considerable amount of work on cosmic structure .
current tests usually exploit the angular correlation function and several harmonic amplitudes of the sky temperature field ( see e.g , adams et al .
1992 ; kashlinsky 1992 ; efstathiou et al . 1992 ; kofmann et al . 1993 ; gorski et al . 1994 ) .
however , several more tests have been suggested over the years for a thorough investigation of the properties of the anisotropy field of the cosmic background radiation .
these often involve the distribution and features of hot and cold spots , which can provide useful checks of the gaussian nature of the fluctuations ( sazhin 1985 ; bond & efstathiou 1987 ; coles & barrow 1987 ; coles 1988 ; martnez - gonzlez & sanz 1989 ; gott et al .
measurable quantities include the number of spots @xmath3 defined by isotemperature contours , the spot boundary curvature or genus @xmath4 , the spot excursion area , and so on .
a first analysis of cobe - dmr maps along these lines has been performed by torres ( 1994 ) .
a definition of spots independent of isotemperature contours considers local maxima and minima of temperature ( bond & efstathiou 1987 , vittorio & juszkiewicz 1987 ) , and is thereby not connected to the topological features of spots .
differing from @xmath5 the dependence of the number of ( positive and/or negative ) peaks @xmath6 on threshold is not universal for gaussian fields .
this latter approach was adopted by fabbri & natale ( 1993 , 1995 ) in studies of the 2-dimensional distribution of extragalactic iras sources , but has not yet been applied to the cosmic background radiation . in this work
we analyze the statistics of local maxima and minima in cobe - dmr 2-year maps .
we found that in this kind of analysis the detector noise must be taken into account very carefully since @xmath6 is sensitive also to high order harmonics where noise dominates ( cf .
fabbri 1992 ) .
however , due to a highly nonlinear dependence of @xmath6 on the harmonic strengths , the presence of structured signals in cobe maps reduces its value below the level measured in pure noise maps . ( the identification of genuine peaks in the radiation temperature is not required at all in our analysis . )
we find that the distributions of positive and negative peaks are mutually consistent , and the results from this statistics agree with those of earlier tests .
therefore , we find no evidence for non - gaussian features in the fluctuations .
more precisely , fitting gaussian power - law models of cosmic structure to the peak distribution we recover a clear anticorrelation between the spectral index @xmath7 and the predicted rms quadrupole @xmath8 ( seljak & bertschinger 1993 ; smoot et al .
1994 ) : we get @xmath9 @xmath10k for @xmath11 and @xmath12 @xmath13k for @xmath14 , where the error bars include uncertainties deriving from the treament of noise as well as cosmic variance . these numbers , altough they agree with previous evaluations of the quadrupole from higher order harmonics , are not consistent with its direct determination providing @xmath15 @xmath10k ( bennet et al .
so the recently discovered discrepancy is confirmed by the properties of the peak distribution , which depend on the harmonic content of the angular distribution up to @xmath16
we analyzed the 2-year 53(a+b ) dmr maps processed with a 2.9@xmath17 smoothing ( wright et al .
1993 ) and dipole subtraction . considering the northern and southern hemispheres separately
, we constructed two pole - centered maps , each containing 12892 pixels , using the coordinate transformation @xmath18 , $ ] @xmath19 with @xmath20 and @xmath21 the galactic coordinates . after masking low galactic latitudes , @xmath22 we were left with 8412 pixels per map .
we then looked for temperature peaks using the algorithms of fabbri & natale ( 1993 , 1995 ) .
table 1 gives the no - threshold numbers of peaks , both actually detected and extrapolated to the entire sky and to the north and south hemispheres ( 2nd and 3rd column , respectively ) .
figure 1 reports the extrapolated numbers vs. a threshold factor @xmath23 .
this is the peak height normalized to the sky rms fluctuation @xmath24 ; for cold spots , @xmath6 gives the number of minima below @xmath25 . for distributions of only positive _ or
_ negative peaks the statistical errors at 1-sigma confidence levels are evaluated as @xmath26 with @xmath27 the unmasked fraction of the sky . within such error limits
, we find no significant difference between the distributions of positive and negative peaks ; this result provides support for the gaussian nature of cosmic perturbations . in the figure we also report the full - sky average number of positive and negative peaks
this average will be compared with theoretical models below , because of the smaller ( by a factor @xmath28 relative error .
llll peak set & @xmath29 maps & @xmath29 maps & @xmath30 maps + & @xmath31 & extrapolated@xmath32 & no mask + hot , north & 50 & @xmath33 & 95 + cold , north & 47 & @xmath34 & 99 + hot , south & 48 & @xmath35 & 105 + cold , south & 53 & @xmath36 & 97 + hot , total & 98 & @xmath37 & 200 + cold , total & 100 & @xmath38 & 196 + hot / cold average & 99 & @xmath39 & 198 + for the interpretation of the above data , we must consider that both cosmic signal and noise contribute to the temperature field @xmath40 .
the expectation values of the numbers of local maxima and minima are determined by the variances @xmath41 . for a gaussian field the full - sky @xmath6
is given by @xmath42 { \rm d}y \right\ } , \end{aligned}\ ] ] where @xmath43 ^{-\frac 12},\ ] ] and the properties of the anisotropy field are summarized by the parameters @xmath44 and @xmath45 ( bond & efstathiou 1987 ; fabbri 1992 ) given by @xmath46 \nonumber \\ & \times & \left [ \sum_\ell \left ( \ell -1\right ) \ell \left ( \ell + 1\right ) \left ( \ell + 2\right ) w_\ell ^2a_\ell ^2\right ] ^{-1},\end{aligned}\ ] ] and @xmath47 ^2\left [ \sum_\ell w_\ell ^2a_\ell ^2\right]^{-1 } \nonumber \\ & \times & \left [ \sum_\ell \left ( \ell -1\right ) \ell \left ( \ell + 1\right ) \left ( \ell + 2\right ) w_\ell ^2a_\ell ^2\right ] ^{-1}.\end{aligned}\ ] ] here @xmath48 are form factors taking into account beam shape and any additional smearing effect .
assuming that signal and noise are uncorrelated , their contributions to @xmath49 add up in quadrature in the ( a+b ) maps , @xmath50 , and we need to determine @xmath51 in an independent way .
this can be achieved by means of the ( a@xmath52b ) maps , which can however be used to directly derive the coefficients @xmath53 rather than the @xmath51 .
figure 2 reports the results of a harmonic best fit up to @xmath54 executed on the entire celestial sphere@xmath55 increasing the number of harmonics , best fits ( as well as direct integration by means of @xmath56 ) tend to overrate large-@xmath57 amplitudes .
we checked for this effect by considering the peak statistics for noise maps .
figure 3 gives the numbers of maxima and minima detected in the full - sky noise maps vs. threshold @xmath58 if using eq.s ( [ npeak]-[gammastar ] ) we generate the peak statistics from a set of 50 harmonics with best - fitted amplitudes , we get a low - threshold excess of about 30 peaks with respect to the data in fig .
2 ; the discrepancy increases with the number of harmonics .
since accurate theoretical calculations of peak statistics require at least @xmath5950 harmonics , we tried to fit an analytic form to a more limited set of @xmath60 .
a satisfactory choice is @xmath61 .\ ] ] however the large error bars on individual @xmath60 make a 2-parameter fit for the function ( [ anoise ] ) not very useful .
a more convenient solution to this problem is to fit the parameter @xmath62 directly on the noise - generated peak distribution ; using the average distribution of maxima and minima ( represented by the filled circles in figure 3 ) and 99 harmonics the result is @xmath63 ( we checked however that a set of @xmath64 harmonics would provide a sufficient accuracy . )
the fit turns out to be even `` too '' good , providing @xmath65 against 9 degrees of freedom : this means that data points at different thresholds are not uncorrelated .
the value found for @xmath66 is quite independent of @xmath67 , since peak statistics does not depend on the overall amplitude of anisotropies as shown by eq.s ( [ tetastar ] , [ gammastar]).we then fitted the amplitude parameter of eq .
( [ anoise ] ) on the reported set of @xmath68 with @xmath69 getting @xmath70 mk@xmath71 the distribution described bu the full line in fig .
3 was generated using the function ( [ anoise ] ) in the entire range @xmath72 we checked that the peak distribution of the ( a - b ) maps is also recovered to a good accuracy building up the harmonic spectrum with the `` measured '' @xmath60 up to @xmath54 and the function ( [ anoise ] ) at higher @xmath73 we tested power - law spectra against the peak statistics in fig .
1 with the following procedure . for each spectrum ,
labelled by the spectral index @xmath7 and the predicted quadrupole @xmath74 we generated the theoretical expectation values @xmath75 and then @xmath76 a_{{\rm s}\ell } ^2+a_{{\rm n}\ell } ^2.\ ] ] here @xmath77 denote the measured beam - shape coefficients of cobe - dmr .
these have reported by wright et al .
( 1994 ) up to @xmath78 a more extensive set of @xmath77 up to @xmath79 has been provided to us by kogut ( private communication ) .
the exponential factor in eq .
( [ totalal ] ) takes into account the 2.9@xmath17 smearing on the map as well as the additional smearing due to orbital motion , so that our best estimate is @xmath80 .
note that @xmath81 should not be confused with the approximate gaussian beamwidth which was used in many computations , but not in the present work . also , it is quite different from the phenomenological parameter @xmath66 of eq .
( [ anoise ] ) .
the peak distributions calculated from eq.s ( [ npeak]-[gammastar ] ) are tested against the averaged distribution of positive and negative peaks .
we should notice that such theoretical distribution are affected by several sources of uncertainty .
cosmic variance ( white et al . 1993 ) affects the cosmic - structure contribution to the harmonic coefficients @xmath82 , and a quite similar variance pertains to noise ; in fact , these two effects are described by identical equations assuming that both are gaussian processes . performing a limited number of simulations for superpositions of gaussian signal and noise ( with fixed @xmath83 , the _ expectation _ values of harmonic strengths )
, we found that the probability distribution of @xmath84 at a given threshold is roughly consistent with a poisson distribution having a width @xmath85 however , in our case the @xmath60 themselves are not fixed .
when we use eq .
( [ anoise ] ) their errors at different @xmath57 are correlated due to the uncertainties on @xmath67 and @xmath86 it can be shown that the corresponding contribution to the uncertainty on the predicted @xmath6 is given by @xmath87 ^2\sigma _ { { \rm n}}^2{\rm var}(\sigma _ { { \rm n}}).\end{aligned}\ ] ] when we use the measured @xmath88 a more familiar equation @xmath89 $ ] applies . for each computed model we calculated the corresponding @xmath90 combining quadratically the above errors with with the experimental error bars reported in fig.1 .
this procedure allowed us to avoid a more extensive use of simulations .
figure 4 gives the allowed regions at 1 and 2 sigmas in the @xmath91 plane .
full lines give the contours calculated for @xmath92 using the @xmath60 of fig .
2 up to @xmath54 , and the analytic form ( [ anoise ] ) with the optimal values of @xmath93 and @xmath67 at higher @xmath73 ( this procedure is referred to as s31 in the figure and in table 2 . )
contours calculated for @xmath94 ( case s29 , not reported in the figure ) would be hardly distinguishable from them . using the analytic form ( [ anoise ] ) in the entire range @xmath95 ( case f31 ) the contours
are slightly displaced to smaller values of @xmath8 ( by @xmath96k for @xmath97 . the parameters @xmath7 and @xmath8 are clearly anticorrelated , as already found from analyses of harmonic amplitudes and the angular correlation function ( seljak & bertschinger 1993 ; torres et al .
1994 ; smoot et a. 1994 ) .
minimizing @xmath90 for fixed @xmath7 we identify a straight line in the @xmath113 plane , @xmath114 with the coefficient values listed in table 2 . from the above results , taking into account differences arising from the s31 and f31 procedures
, we can conclude that @xmath9 @xmath10k for @xmath11 , and @xmath115 @xmath10k for @xmath14 .
these numbers agree very well with the most likely quadrupole @xmath8 derived from higher order multipoles , but not with the quadrupole rms fluctuation of 6@xmath116 @xmath10k directly fitted on two - year data ; see bennet et al .
( 1994 ) for a discussion of this discrepancy , and jing & fang ( 1994 ) for a possible explanation in terms of an infrared cutoff in the spectrum .
we wish to thank a. kogut for providing unpublished data on cobe - dmr beam shape .
this work is partially supported by agenzia spaziale italiana under contract # 94-rs-155 , by the italian ministry for university and scientific and technological research ( progetti nazionali e di rilevante interesse per la scienza ) , and by the european union under contract # ci1 * -ct92 - 0013 . | we perform the stastistics of temperature maxima and minima in cobe - dmr 2-year maps . for power - law
spectra the surface distribution of peaks implies an amplitude consistent with more conventional analyses of cobe data ( for instance , we get @xmath0k for a spectral index @xmath1 but not with the measured quadrupole @xmath2k .
this provides further support for the existence an infrared cutoff in the cosmic spectrum . |
in the nonrelativistic constituent quark model for mesons and baryons the quarks are bound by confining potentials . despite their limitations and concerns about their validity
, these potential models describe the properties of the various mesons and baryons surprisingly well ( see , for example , silvestre - brac @xcite and the two excellent review articles by lucha et al .
@xcite and richard @xcite on the matter ) .
once the potential model approach is adopted , the three - quark problem can be solved using various approaches . among them
, the hyperspherical harmonics ( hh ) method is quite successful in applications as it is well suited to describe radial and orbital excitations @xcite . within the framework of the hh method
, the hypercentral approximation ( hca ) has been used in the past @xcite to study the spectrum of the baryon .
there are various reasons for adopting the hca to study the three quark system : i ) the two - body potential acting between quarks is quite soft and therefore in the hh expansion of the interaction only the first term gives a significant contribution to the binding energy .
this of course means that the two - body correlations are not as strong as compared to the nuclear correlations ; ii ) it is quite simple and thus one avoids the complicated three - body calculations via , for example , the faddeev equations @xcite , and iii ) the results obtained from it are accurate and the spectra are well reproduced .
another method , in the framework of the hh method , is the integrodifferential equation approach ( idea ) @xcite which includes higher terms of the hh expansion in an average manner .
the idea method takes two - body correlations into account exactly , reproduces the spectrum of the nucleon quite well , and provides wave functions reliably @xcite which is crucial in studying photoexcitation processes .
these processes are manifested as resonances and can be excited through electromagnetic transitions giving rise to large enhancements in the total absorption cross section @xcite .
the photoexcitation of the nucleon resonances has been studied in the past by various groups @xcite .
the results obtained by them are rather unsatisfactory when compared to the experimental data .
the inclusion of retardation effects and relativistic corrections does not improve the situation much @xcite . in this work
we consider the absorption of a single photon by a nucleon which then undergoes a transition from the ground state to an excited one .
the photoabsorption cross section is calculated using various quark - quark potentials and by using the hca and idea methods . in sec .
2 we describe our formalism . in sec .
3 we give details on how the @xmath0 and @xmath1 transition amplitudes are calculated while in sec .
4 we present our results and discussions .
the photoexcitation process is described by the transition amplitude @xmath2 where @xmath3 is the initial ground state wave function of the nucleon , @xmath4 is the wave function of the final excited state , and @xmath5 the perturbative electromagnetic hamiltonian . in what follows
we shall discuss these ingredients in some detail .
the fully antisymmetric total wave function for a three - quark system can be expressed as a product of configuration space , flavor , spin , and color functions .
since baryons are color singlets , the color wave function is totally antisymmetric ( @xmath6 ) and thus the remaining product must be fully symmetric ( @xmath7 ) , @xmath8 the structure of the symmetric component of eq .
( [ psi ] ) depends on the transition considered and can be constructed using the various symmetries involved .
+ for the construction of the symmetric part of the total wave function the fully symmetric , mixed symmetric , and mixed antisymmetric configuration space wave functions are required .
these can be obtained using the idea @xcite method . in this method
the fully symmetric ground state configuration space wave function is constructed from the faddeev - type components @xmath9 @xcite @xmath10\ , , \label{ideas}\ ] ] where ( @xmath11 ) are the jacobi coordinates , @xmath12^{1/2 } $ ] is the hyperradius with @xmath13 , and the @xmath14 are given by @xmath15 the required mixed symmetry states for @xmath16 are given by @xmath17 \bigg \ } \,,\nonumber \\
\psi_1^{m^a}(\vec{\rho},\vec{\sigma } ) & = & \frac{1}{r^{5/2 } } \bigg [ ( 1+z_{31})^{1/2}y_{10}(\omega_{31})p_1^{s^\prime}(z_{31},r ) \nonumber \\ & - & ( 1+z_{23})^{1/2}y_{10}(\omega_{23})p_1^{s^\prime}(z_{23},r ) \bigg ] \,,\end{aligned}\ ] ] where the superscripts @xmath18 and @xmath19 denote the mixed symmetric and antisymmetric states with respect to the interchange of particles 1 and 2 .
the required symmetric spin - flavor states are given by @xmath20\,,\ ] ] while the mixed symmetric states are @xmath21\,,\\ \left |\xi^{m^a } \right > & = & \frac{1}{\sqrt{2 } } \left [ \phi^{m^s } \chi^{m^a } + \phi^{m^a } \chi^{m^s } \right ] \,.\end{aligned}\ ] ] the relevant flavor and spin states are given by various authors and therefore , will not be presented here ( see , for example , refs .
the singlet , antisymmetric color state , @xmath22 where r , b and y stand for red , blue , and yellow respectively , does not enter into the calculations and therefore , in what follows will be suppressed .
the initial total wave function for the proton ( p ) ground state , with @xmath23 , and @xmath24 , is given by @xmath25 \left|\psi^s_{0}\right>\ , , \label{protong}\ ] ] where the lower index of the space wave function @xmath26 refers to the angular momentum @xmath27 .
the final wave function for the first excited state , with @xmath28 , and @xmath29 , of the proton is @xmath30\,.\ ] ] for the @xmath1 transition @xmath31 , where the proton and @xmath32 both have an angular momentum @xmath33 , the total wave function for the initial state of the proton is still given by eq .
( [ protong ] ) , while the final wave function for the delta is @xmath34 the perturbative hamiltonian for the electric dipole e1 transition , in the case of three quarks of equal mass m , is given by @xmath35 where @xmath36 is the momentum of quark @xmath37 and @xmath38 denotes a polarization direction of the incident photon . for @xmath39 and @xmath40 quarks , the charge operator has the form @xmath41 where @xmath42 is the third component of the isospin of the @xmath37-th quark with @xmath43 using commutation relations to express the momenta in terms of the three - quark hamiltonian @xcite we may rewrite ( [ he11 ] ) in a siegert form @xcite @xmath44 where @xmath45 is the coordinate of @xmath37-th quark conjugate to the momentum @xmath46 . in jacobi
coordinates we have @xmath47 where @xmath48 is the difference between the final and initial binding energies and the operators @xmath49 and @xmath50 are given by @xmath51 thus , instead of expressing the hamiltonian in terms of the individual particle charge and coordinates , the more appropriate jacobi coordinates and operators @xmath52 and @xmath50 which act on quasi - particles @xcite are used .
the magnetic dipole @xmath53 causes the transition @xmath54 , in which the proton , after absorbing a photon ( @xmath55 ) , is excited to the delta ( @xmath56 ) .
the corresponding perturbative hamiltonian is @xmath57 where @xmath58 is the magnetic moment operator of the @xmath37-th quark .
since @xmath59 does not contain any orbital operators , in this transition the spin must change instead .
noting that @xmath60 and by letting the charge operators @xmath52 and @xmath50 act on the three - quark isospin states we finally obtain for the transition matrix elements @xmath61 the integrals in ( [ me11 ] ) were evaluated using euler angles @xmath62 as external and @xmath63 as internal coordinates . here
the @xmath64 axis is chosen to coincide with @xmath65 and @xmath66 is lying in the @xmath67 plane .
thus only a five dimensional integration has to be done numerically since both @xmath68 and the two components of @xmath69 are invariant with respect to a rotation about the z - axis and thus do _ not _ depend on @xmath70 . after averaging over the direction of @xmath71 and polarization direction
one obtains the following expression for the absolute square of the transition matrix elements @xmath72 therefore , the following integrals are required @xmath73 the corresponding integrated photoabsorption cross section for a single excited electric dipole state , in the long wavelength limit @xcite , is @xmath74 for the @xmath1 transition amplitude we have @xmath75 where @xmath3 and @xmath4 are given by eqs .
( [ protong ] ) and ( [ fwp ] ) respectively . the process @xmath76
( @xmath77 ) can take place either via the magnetic dipole ( @xmath1 ) or the electric quadrupole ( @xmath78 ) transition . in the quark model the latter transition is forbidden @xcite because it is proportional to the charge operator which can not cause transitions between quark spin @xmath79 and @xmath80 states , and hence the matrix element vanishes by orthogonality of the quark spin wave functions .
the @xmath1 transition involves the quark magnetic moments hence the spin operator and this can lead to transitions @xmath31 @xcite . the transition matrix element can be written as @xmath81 using the the flavor , spin , and configuration space wave functions and averaging over the two photon polarization directions we finally obtain @xmath82 where @xmath83 is the overlap integral given by @xmath84 like in the electric transitions , the photoabsorption cross section for a single excited magnetic dipole state is @xmath85
the quark - quark potential can be written as a sum of the central and the spin - spin parts @xmath86 where @xmath87 contains , as usual , the confinement and the coulombic parts while @xmath88 is of the general form @xmath89 the spin - dependent interaction , suggested by the perturbative one - gluon exchange mechanism , is important in describing the splitting of the meson masses and the experimental mass difference @xmath90 .
it contains the function @xmath91 which is singular as @xmath92 and therefore the corresponding wave equation has no physically acceptable solutions and one expects a collapse in both the quark - quark and in baryon systems . to avoid this a cut - off or a smearing function is introduced to reduce the singularity and to make the calculations tractable .
since the resulting spin - spin potential is strongly attractive and short - ranged , it can generate significant short range correlations which have stronger effects on the ground state than on the excited @xmath93 states .
the strong correlations at short distances in the @xmath33 partial wave are not expected to play a major role in photoexcitation processes as the excited states are shifted to the outer region and thus in the overlap integral short - range contributions are rather unimportant
. therefore any influence of the spin - spin force will come from the modification of the mass differences and of the corresponding wave functions . in this work we performed calculations with the @xmath87 term alone as well as with both terms
for the calculations with the @xmath87 part only , we employed the martin @xcite , the cornell @xcite , and the lichtenberg @xcite potentials . for the calculations
in which both terms are included we used the ono - schberl potential @xcite and two of the recently published potentials by silvestre - brac @xcite , namely , the ap1 and ap2 versions which have the general form @xmath94 with @xmath95 the parameters for ap1 are given by @xmath96 while those of ap2 are given by @xmath97 we remind here the reader that the quark - quark potential @xmath98 is related to the quark - antiquark potential @xmath99 by lipkin s rule @xcite , i.e. , @xmath100 .
the results obtained for the ground state and the first orbitally excited state of the nucleon with spin - independent potentials and by using the idea and hca methods are given in table i. both methods are in very good agreement with each other and the transition energy @xmath101 , shown in table ii , is reasonably well reproduced .
this transition energy and the corresponding wave functions were used to calculate the integrated photoabsorption cross section @xmath102 due to the e1 transition , from the ground state to the n(1520 ) resonance .
our results are given in table ii together with experimental values and those of other methods .
it is seen that for both the idea and the hca the cross sections are in good to very good agreement with the experimental data .
the results obtained by other methods for the transition energy are generally very low as compared to the experimental ones except those obtained via the isgur - karl ( i.k . )
model @xcite , which , nevertheless , can not reproduce the experimental photoabsorption cross section @xmath102 well .
this mainly implies that the corresponding wave functions are not adequate to describe the photoabsorption cross section for the e1 transition .
the latter is also true for the wave functions employed by brizzolara and giannini in their thorough investigations on the nucleon photoabsorption @xcite based on the nonrelativistic quark model of isgur and karl and the bag model .
they explicitly demonstrated that the results are rather strongly dependent on the hyperfine mixing and on the dimensions of the system and thus on the h.o parameter used .
the discrepancy with the experimental results still persists even when other contributions such as charge plus current densities with and without retardation effects are incorporated into the h.o model .
three - body force components of the quark interaction ( in the 3q model ) , and relativistic corrections in the framework of the bag model do not improve matters to a significant degree either .
apart , of course , from the inadequacy of the wave functions these discrepancies probably stem also from some other fundamental absorption mechanism , such as the coupling to mesons and/or @xmath103 pairs @xcite . the results for the nucleon masses obtained by using spin - dependent potentials are given in table iii .
the corresponding transition energy @xmath101 together with the cross sections are presented in table iv .
it is seen that both ap1 and ap2 potentials of silvestre - brac give excellent results for the photoabsorption cross section @xmath102 .
this came as a surprise since the transition energies are not as good as one would like them to be .
this can be attributed to the fact that in the confining potential the excited state lies at a higher energy as compared to the experimental value and thus the wave function for @xmath16 is more spread out in space and therefore the overlap integral somehow compensates for the excess transition energy .
our results obtained with the ono - schberl potential are in fair agreement with the experimental data .
the delta masses are given in table v while the transition energy @xmath104 is shown in table vi ; it is well reproduced but the results for the corresponding integrated photoabsorption cross section of the @xmath105(1232 ) resonance are only in fair agreement with the experimental values .
the same is true for the i.k model .
this discrepancy can be connected to the strong dependence on the quark masses that enter the cross section via the magnetic moments .
finally the energy and cross section results obtained via the idea and hca methods are quite similar .
this comes as no major surprise as the quark - quark potentials are soft and thus two - body correlations do not play a significant role .
this means that the corresponding wave functions are fairly similar and the small differences between them do not strongly influence the overlap integral and thus the cross sections . in the case where a spin - spin force was used the short range correlations
are not manifested in the overlap integral either due to the shifting of the @xmath106 wave function to the outer region .
in conclusion the nonrelativistic potential model reproduced the experimental data for the e1 transitions quite successfully .
there is a weak potential dependence and in the case of the spin - independent potentials the best results are obtained with the lichtenberg potential . in the case of the spin - dependent potentials both ap1 and ap2 give excellent results for the e1 photoabsorption cross sections . in the m1 transitions where there is a strong dependence on the quark
masses the ono - schberl potential gives better results than the ap1 and ap2 potentials .
; . in , vol . 273 of , edited by ( springer - verlag , new york , 1986 ) , pp .
283 - 323 . ; ; . .
; . ; . , academic press , 1979
. , new york , john wiley and sons 1967 . . . . ,
amsterdam , north holland 1970 . . . .
1.2cmcccccccccc & quantum & & & & idea & & & hca & + & numbers & & expt & & potentials & & & potentials & + l & s & @xmath107 & values & m & c & l & m & c & l + 0 & 1/2 & @xmath108 & 940 & 1085 & 1087 & 1085 & 1090 & 1090 & 1088 + 1 & 1/2 & @xmath109 & 1520 & 1676 & 1773 & 1730 & 1676 & 1774 & 1730 + 2mmcccccccc & & & idea & & & hca & + & expt & & potentials & & & potentials & + & values & m & c & l & m & c & l + @xmath101 & 580 & 591 & 686 & 645 & 586 & 684 & 642 + ( mev ) & & & & & & & + @xmath102 & @xmath110 & 30.4 & 27.8 & 30.6 & 30.6 & 28.6 & 31.3 + ( mevmb ) & & & & & & & + & & & other & methods & & & + & & i.k . & h.o . & h.o . & h.o . & 3q & bag + & & & & ( ret ) & ( @xmath111 ) & & + @xmath101 & 580 & @xmath112 & 167 & 167 & 167 & 144 & 349 + ( mev ) & & & & & & & + @xmath102 & @xmath110 & @xmath113 & @xmath114 & 38 & 43 & 27 & 13 + ( mevmb ) & & & & & & & + 2mmcccccccccc & quantum & & & & idea & & & hca & + & numbers & & expt & & potentials & & & potentials & + l & s & @xmath107 & values & ap1 & ap2 & os & ap1 & ap2 & os + 0 & 1/2 & @xmath108 & 940 & 1028 & 1021 & 920 & 1035 & 1029 & 940 + 1 & 1/2 & @xmath109 & 1520 & 1772 & 1797 & 1821 & 1773 & 1797 & 1821 + 2mmcccccccc & & & idea & & & hca & + & expt & & potentials & & & potentials & + & values & ap1 & ap2 & os & ap1 & ap2 & os + @xmath101 & 580 & 744 & 776 & 901 & 738 & 768 & 881 + ( mev ) & & & & & & & + @xmath102 & @xmath110 & 32.6 & 32.3 & 27.1 & 32.9 & 32.5 & 27.4 + ( mevmb ) & & & & & & & + 2mmcccccccccc & quantum & & & & idea & & & hca & + & numbers & & expt & & potentials & & & potentials & + l & s & @xmath107 & values & ap1 & ap2 & os & ap1 & ap2 & os + 0 & 3/2 & @xmath120 & 1232 & 1300 & 1307 & 1224 & 1300 & 1308 & 1225 + 2mmccccccccc & & & idea & & & hca & & other methods + & expt & & potentials & & & potentials & & + & values & ap1 & ap2 & os & ap1 & ap2 & os & i.k .
+ @xmath121 & 292 & 272 & 286 & 304 & 265 & 279 & 285 & @xmath122 + ( mev ) & & & & & & & & + @xmath123 & @xmath124 & 81 & 83 & 54 & 79 & 82 & 58 & 52 + ( mevmb ) & & & & & & + | photoexcitation of the nucleon is studied in the framework of the constituent quark model with nonrelativistic quark - quark interactions . utilizing the dominance of the pairwise structure of the relevant three - quark potential ,
the bound and excited state wave functions have been constructed using the faddeev - type integrodifferential equation approach .
the method is based on the hyperspherical harmonics expansion and takes two - body correlations exactly into account and therefore reliable wave functions , required in photo processes , can be obtained .
the integrated photoabsorption cross sections for the @xmath0 and @xmath1 transitions are calculated and compared with experimental values as well as with those of other theoretical calculations .
our results , in overall , are in good to very good agreement with the experimental values . |
we present a study on the performances of a gravitational wave ( gw ) observatory composed by a hybrid network of detectors .
in particular , we focus on the possibility to use a resonant detector to perform gw observations with one interferometric detector .
this could be an opportunity in the scenario after ligo s5 run and the first virgo science run , when most of the interferometers will be shut down for upgrading : current plans are that geo will be kept in operation till the start of the ligo s6 and the second virgo science runs , supported by ligo hanford 2k detector over weekends . in this sense , we present a case study on joint observations between auriga and virgo on a test period of 24 hrs . in the past years
, various searches for gw signals have been independently performed by networks of resonant bars @xcite or interferometers @xcite .
there have been also some attempts to perform burst searches among detectors with different spectral sensitivity and orientation : by tama and the ligo scientific collaboration ( lsc ) @xcite , by auriga and the lsc @xcite and by the infn bars and the virgo collaboration @xcite .
the proposed network search strategy takes as a starting point the waveburst+corrpower @xcite search used by lsc for the s3 and s4 analyses @xcite .
that search was greatly innovative : a two - step search composed of an excesspower - like event trigger generator plus a cross - correlation test which allowed an efficient reduction of false alarms .
in that case , however , the detectors partecipating to the network were almost aligned and had a similar spectral sensitivity .
an extension of such methodology to the case of similar but misaligned detectors has been discussed in literature @xcite .
the novelty of our work consists in a further generalization to detectors with different spectral sensitivities , so that it can be implemented between a resonant bar and an interferometer . to better characterize the method
, we compare its performances with those of a simple time - frequency coincidence search .
the paper is organized as follows : in section 2 we introduce the search method .
section 3 presents an overview of the exchanged data and summarizes the main steps of the network pipeline and of the tuning performed on chosen test - statistics .
results and conclusions are presented in section 4 and 5 respectively .
the gw search method described in this paper is characterized by two main parts : the event trigger generator , whose role is to select a subset of _ interesting _ triggers and a coherent analysis .
the trigger search is based on waveburst @xcite , an excess power algorithm based on the wavelet decomposition in the time - frequency plane . in the present work
, waveburst has been used in coincidence mode , i.e. the algorithm selects time - frequency coincident excesses of power between the two detectors .
the step of coherent analysis is based on a cross - correlation test between data streams weighted by a combination of the strain sensitivities of the two detectors ( xcorr ) .
[ method ] the strain produced on the detector @xmath4 by an incoming burst signal with polarization components in the wavefront frame @xmath5 is : @xmath6 where @xmath7 and @xmath8 are the _ antenna pattern _
functions @xcite ( and references therein ) . following @xcite , eq.[gen_h ]
becomes : @xmath9\ ] ] where @xmath10 is a directional sensitivity , @xmath11 is a phase shift and ( @xmath12 ) is the location of the gw source . and @xmath13 $ ] . ]
the reconstructed strain at the input of two detectors , @xmath4 and @xmath14 , is @xmath15 , where @xmath16 are the two independent noises .
it has been shown in @xcite that the following linear combination of the two reconstructed strains , called _ null stream _ , cancels the signal : @xmath17 , where @xmath18 includes the light travel time and a suitable fraction of the typical gw period , so that @xmath19 . whenever one considers a cross - correlation with a narrow - band detector , this approximation is automatically verified since the reconstructed strain at input has to be bandlimited by a suitable filter ( see fig.[xcorr - filter ] ) . ] .
we introduce the fourier transform of the null stream and normalize it to its variance per unit bandwidth , @xmath20 : we properly filter @xmath21 obtaining the _ weighted _ reconstructed strain at detector @xmath4 : @xmath22 where @xmath23 are the noise power spectral densities of the detectors in terms of gw strain .
hence , the normalized null stream is @xmath24 .
one well known method to search for gw signals in the data relies on the minimization of the null energy @xcite , @xmath25 , where the time integral is performed on the signal duration plus any typical response time of the narrower band detector .
the null energy can be expressed in terms of the correlated energy and the incoherent energy of the network : @xmath26 .
the former is the contribution of the cross - correlation of the detectors , @xmath27 .
the latter is the auto - correlation contribution of the detectors , @xmath28 .
as discussed in @xcite , a gw candidate is selected against background events more efficiently by maximizing @xmath29 rather than by minimizing @xmath30 .
in fact , @xmath30 can take low values even for accidental events with small @xmath29 and @xmath31 ; instead , for detectable gw signals , we expect a higher @xmath31 , almost balanced by a positive @xmath29 . for these reasons
, this coherent step of network analysis is based on the maximization of the correlated energy @xmath29 in our null stream . in principle
, @xmath29 depends on @xmath12 and @xmath3 of the source through @xmath18 , @xmath10 and @xmath32 .
however , we checked that in the case of random polarized gw radiation , emitted by sources distributed either in the galaxy or uniformly in the sky , we can follow an approximated maximization procedure of @xmath29 assuming @xmath33 .
the main reason is that auriga is limiting the common bandwidth and at the same time shows a worse strain power spectral density ; therefore , the @xmath34 defined in eq.[cross - freq ] is dominated by the strain noise of auriga unless @xmath35 .
this condition however is occurring rarely enough : @xmath36 with 90 % probability for a galactic population of sources , and very similar values occur also considering the ligos or geo interferometers in place of virgo .
fig.[xcorr - filter ] shows the @xmath34 , corresponding to @xmath37 .
these curves are quite similar and therefore it is advisable to approximate @xmath38 in eq.[cross - freq ] , i.e. to consider filtered data streams : @xmath39 to compute the cross - correlation . in addition , to be robust against noise fluctuations and relative calibration errors between auriga and virgo , we considered the r - statistics of such filtered data streams : @xmath40 where we integrate over a time window of fixed duration and search for local maxima as a function of the time @xmath41 and time shift @xmath18 between the data streams .
we analysed the 24 hours of data starting from the utc time 14 sep 2005 23:11:27 . in this time period
, virgo was performing its seventh commissioning run ( c7 ) while auriga was in stable operation .
this affects mainly the quality of the data of virgo for the presence of periods of instrumental unlock reducing the observation time from 24 to 16 hours and 33 mins .
moreover , the virgo collaboration provided a list of vetoes to flag the most noisy periods , corresponding to @xmath42 reduction of the live time .
the @xmath43 sensitivity curves for the two detectors in the @xmath44 $ ] hz band on september @xmath45 are shown in fig.[xcorr - filter ] .
fig.[pipeline ] shows the scheme of the implemented pipeline : as a first step , the raw data from the two detectors are whitened and bandpassed to match the auriga bandwidth ( i.e. 850 - 950 hz ) , then they are fed in as input to waveburst @xcite .
waveburst has been configured to search for coincident excesses of power on the two data sets over 4 wavelet decomposition levels , corresponding to time / frequency resolutions ranging from 4.916 ms by 101.72 hz to 39.32 ms by 12.72 hz ; it then produces the list of triggers , containing all the trigger parameters ( e.g. the central time in gps seconds , the central frequency , the frequency range , the geometric significance , @xmath46 , the signal to noise ratio , snr , etc . )
; for each of those events , the last step of the pipeline , xcorr , selects the corresponding chuncks of weighted filtered data ( see sec .
[ method ] ) and , by sliding on @xmath41 and @xmath18 the two time series , ms and the phase shift is @xmath47 ms ; to be conservative we set the relative time slide to @xmath48 ms . ]
it searches the maximum of the cross - correlation coefficient , @xmath49 , computed over a given time window . within our approach ,
@xmath46 and @xmath50 are the main test statistics which characterize each event at the output of the pipeline .
a combined threshold on these quantities sets the overall false alarm rate ( far ) and detection efficiency of the network .
other internal parameters have been set in order to have neglegible impact on the final far and detection efficiency .
the background is estimated by applying the pipeline to 4000 time - shifted data sets with a total live - time of 2760 days , with shifts ranging from -2123s to + 2123s in steps of @xmath51 s. this choice allows to have a sufficiently large number of resamplings , while preserving the main characteristics of the data sets .
the detection efficiency has been computed over a number of different waveforms ( mainly elliptically polarized damped sinusoids ) by means of software injections : the signal central frequency @xmath52 ranges within the bar bandwidth ( 850 - 950 hz ) and the decay time @xmath53 spans at most a few tens of milliseconds , with random inclination and polarization angles .
we have considered source populations distributed at the galactic center , in our galaxy according to its mass distribution @xcite and uniform distributed in the sky . for each waveform ,
we have created two sets of 994 signals , approximately spaced by 60 s : the _ tuning _ and the _ estimation _ sets .
the aim of the tuning phase is to optimize the pipeline parameters in order to achieve a low far , while preserving the detection efficiency .
the search pipeline has been tuned using 2000 time - shifts randomly selected out of the initial 4000 and the _ tuning _ set of injections .
the parameters set in this tuning phase have then been used on the second half of the time shifts to get the final far and on the _ estimation _ set of injections to calculate the network efficiency . [
[ cross - correlation - window ] ] cross - correlation window + + + + + + + + + + + + + + + + + + + + + + + + among the pipeline parameters , the cross - correlation window is one of the most important .
the general view is that a short time window may be unsuitable for a long signal , as a significant part of its power may be cut off by the window itself . at the same time , if the integration time span is too large , the signal may be diluted in it .
indeed , the weighted cross - correlation filter induces a correlation between the 2 data series over a time scale of the order @xmath54ms .
therefore , even for delta - like signals , we need a cross - correlation window of at least @xmath55 ms . we have set it to 100 ms after testing the performances of various time windows ( [ 25 , 50 , 100 , 200 , 400 , 800]ms ) in terms of far and efficiency for narrow and wide band simulated signals . [
[ combined - threshold ] ] combined threshold + + + + + + + + + + + + + + + + + + we set the far to @xmath56 : given our short observation time , we could not target a smaller rate , since only 7 background events were surviving .
we have not applied the virgo list of vetoes , as from tests performed on the tuning set of the shifts , we measured a @xmath57 far reduction only , while losing @xmath42 of the live time .
the target far has been achieved by setting a combined ( @xmath58 ) threshold on the event list selected by the pipeline . in fig.[eff 930 - 30 , 914 - 14 ] , we show an example of how a population of injected signals ( squares ) detected by the pipeline is distributed with respect to the background events ( dots ) for two waveforms .
the dashed line shows the optimized combined threshold on ( @xmath59 ) : such threshold is defined by a simple linear equation : @xmath60 , where the parameters @xmath61 are set in such a way as to achieve the maximum efficiency at our target far .
we underline here that , since auriga performance over the short time period of the selected data is quite stable , its noise distribution does not change significantly over different time spans and hence the choice of our threshold is fairly general . [
cols="^,^ " , ] we have compared the results of this analysis ( i.e. detection efficiency , far , observation time ) with those of a simple time - frequency coincidence of single detector triggers . in the latter pipeline
, we have analysed the two data streams separately using waveburst as event trigger generator and the resulting triggers retained if coincident in time ( @xmath62 ms ) and frequency ( @xmath63 ) .
we have then applied the list of vetoes for virgo c7 data to the surviving triggers , reducing by a factor @xmath64 the accidental coincidences at the cost of @xmath42 of live time .
finally , an optimized threshold on snr has been applied on the resulting events from the tuning set , in order to get the target far@xmath65 .
the best efficiency has been achieved by setting the threshold at @xmath66 for auriga and @xmath67 for virgo . on the estimation sets , the far results @xmath68 and the corresponding detection efficiencies are reported on in fig.[effgc - gd - as].b for a galactic center source population . at similar
far , the proposed pipeline outperforms the time - frequency coincidence for the tested waveforms in detection efficiency .
we investigated the performances of a joint search for gw bursts by a narrow - band resonant detector and a wide - band interferometer using one day of data taken by auriga and virgo in september 2005 .
the data analysis method based on cross - correlation outperforms a simpler time - frequency coincidence search : it achieves better efficiencies at equal false alarm rate , mostly because it is by far more powerful in discriminating accidental coincidences .
this search monitors at 50% efficiency galactic sources emitting @xmath69 in gw bursts of @xmath70 ms decay time , provided that their strongest fourier components are in the auriga bandwidth .
the resulting detection efficiency of the hybrid network is limited by the less sensitive detector , auriga , to @xmath71 times larger amplitudes with respect to a virgo only search on a larger c7 data set @xcite for pulses of @xmath72 duration . on the other hand
, there is a clear advantage of this hybrid search in the background reduction with respect to a single detector search , as it allows to identify candidate events with high statistical confidence .
the main limitations of this methodological study comes from the short duration of the data set used , which prevented us from investigating false alarm rates lower than a few per year .
we expect that the efficiency of this methodology would take only a small benefit from the achieved progresses in sensitivity of the interferometric detectors virgo , ligos and geo ; however , these progresses should reduce significantly the false alarm rate of this search
. further studies with newer and longer data sets are necessary to assess quantitatively this issue .
nevertheless , this hybrid network search could be of interest for the near future , when only one interferometer will be taking data , a likely condition for a large fraction of the time due to the planned instrumental upgrades towards enhanced detectors .
10 url # 1#1urlprefix[2][]#2 allen z _ et al . _ 2000 _ physical review letters _ * 85 * 50465050 | we present a method to search for transient gws using a network of detectors with different spectral and directional sensitivities : the interferometer virgo and the bar detector auriga .
the data analysis method is based on the measurements of the correlated energy in the network by means of a weighted cross - correlation . to limit the computational load ,
this coherent analysis step is performed around time - frequency coincident triggers selected by an excess power event trigger generator tuned at low thresholds .
the final selection of gw candidates is performed by a combined cut on the correlated energy and on the significance as measured by the event trigger generator .
the method has been tested on one day of data of auriga and virgo during september 2005 .
the outcomes are compared to the results of a stand - alone time - frequency coincidence search .
we discuss the advantages and the limits of this approach , in view of a possible future joint search between auriga and one interferometric detector . |
there has long been considerable circumstantial evidence that at least some luminous , low - redshift qsos are the result of strong interactions or mergers of galaxies ( _ e.g. , _ * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; see * ? ? ?
* for a review ) .
however , a concrete suggestion for an evolutionary scenario for such objects was lacking until @xcite showed that ultraluminous infrared galaxies ( uligs ) , virtually all of which are compelling examples of ongoing mergers , had bolometric luminosities and space densities similar to those of qsos .
these similarities suggested the possibility that uligs are dust - enshrouded qsos which , after blowing away the dust , become classical qsos . if this hypothesis is correct , one should be able to observe examples of objects that are at intermediate stages of this evolutionary sequence .
we are conducting a study of a sample of low - redshift objects that may be in such a transitionary state .
these objects are recognized as bona - fide qsos and are found at an intermediate position in a far infrared ( fir ) color - color diagram between the regions occupied by typical qsos and uligs ( see fig .
[ firplot ] ) . fir color
color diagrams have been used as tools to detect and discriminate different types of activity in the nuclear and circumnuclear regions of galaxies .
different kinds of objects such as qso / seyfert , starbursts , and powerful ir galaxies , occupy fairly well defined regions in the diagram ( see , _
e.g. , _ * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ) . with deep imaging and spectroscopic observations of the host galaxies
, we are attempting to construct interaction histories for each of these `` transition '' objects .
if strong interactions triggered the qso activity and induced starbursts , one might expect both events to occur roughly simultaneously , since both are plausibly dependent on gas flows to the inner regions .
thus , we are placing these objects on an age sequence by measuring the time elapsed since the last major starburst event . this age sequence along with interaction histories can help us answer the question of whether the intermediate position of these objects is indicative of evolution from the ulig to the classical qso population , or whether it simply indicates a range of characteristics in qsos .
our sample is drawn from the @xcite , @xcite , and @xcite samples of _ infrared astronomical satellite _ ( _ iras _ ) objects , and it consists of those objects which have : ( 1 ) a luminosity above the cutoff defined for quasars by @xcite , _
@xmath3 for @xmath4 kms@xmath5 mpc@xmath5 ( or @xmath6 for @xmath7 kms@xmath5 mpc@xmath5 ) , ( 2 ) a redshift @xmath8 , ( 3 ) a declination @xmath9 , ( 4 ) firm _ iras _ detections at @xmath10 m , @xmath11 m , and @xmath12 m , and ( 5 ) a position in the fir color color diagram which is intermediate between the ulig and qso loci ( fig . [ firplot ] ) .
although mrk231 just misses the luminosity threshold given above , its active nucleus is known to suffer heavy extinction ( see [ mrk231 ] ) , apart from which it would clearly be a member of the sample .
we know of no other objects satisfying the other criteria for which this is true .
we have therefore chosen to include it for the present , although it may be approriate to exclude it from some of the analyses of the whole sample , which will be presented in a subsequent paper .
so far , we have presented results for two of the nine objects in the sample : 3c48 ( * ? ? ? * hereafter cs2000 ) , an ongoing merger near the peak of starburst activity ; and pg1700 + 518 ( * ? ? ?
* , hereafter cs97 and scc98 ; see also ; * ? ? ?
* ) , where a tidally disturbed companion with a dominant 85 myr old post - starburst population may be in the process of merging with the host galaxy .
in this paper we present the results for three additional objects : mrk1014 , iras07598 + 651 , and mrk231 .
we assume @xmath7 km s@xmath5 mpc@xmath5 and @xmath13 throughout this paper , so that the projected physical length subtended by 1 is 2.43 kpc for mrk1014 , 2.26 kpc for iras07598 + 651 , and 0.77 kpc for mrk231 .
spectroscopic observations for the three objects were carried out using the low - resolution imaging spectrometer ( lris ; * ? ? ?
* ) on the keck ii telescope . for iras07598 + 6508
, we used a 600 groove mm@xmath5 grating blazed at 5000 yielding a dispersion of 1.28 pixel@xmath5 . for mrk231 and mrk1014
, we used a 300 groove mm@xmath5 grating blazed at 5000 with a dispersion of 2.44 pixel@xmath5 .
the slit was 1 wide , projecting to @xmath145 pixels on the tektronix 2048@xmath152048 ccd .
we obtained two or three exposures for each slit position , dithering along the slit between exposures .
table [ journal ] shows a complete journal of observations , with specification of the slit positions , and total integration times .
the spectra were reduced with iraf , using standard reduction procedures . after subtracting bias , dividing by a normalized halogen lamp flat - field frame and removing sky lines
, we rectified the two - dimensional spectra and placed them on a wavelength scale using the least - mean - squares fit of cubic spline segments to identified lines in a hg - kr - ne lamp .
we calibrated the spectra using spectrophotometric standards from @xcite observed with the slit at the parallactic angle .
the distortions in the spatial coordinate were removed with the iraf _ apextract _ routines . for each slit position , we had two or three individual frames ; we averaged the spatially corrected spectra using the iraf task _
scombine_. we then corrected the spectra for galactic extinction , using the values given by @xcite . since we were aiming to observe the youngest populations in the host galaxies of these objects , we chose the slit positions based on previously obtained color maps of the host galaxies .
we obtained imaging data for the three objects with the university of hawaii 2.2 m telescope as specified in table [ imaging ] .
observations with the @xmath16 filter ( centered at 3410 with a bandpass of 320 ) sample the spectral energy distribution ( sed ) of galaxies on the short side of both the 4000 break and the balmer limit .
this region is very sensitive to the age of the stellar population , and will be brightest in regions of very recent star formation ( ages @xmath17 myr ) , as well as regions with scattered qso continuum .
observations in @xmath18band sample the region redwards of the balmer limit , where @xmath14a - type stellar populations are expected to peak , while longer wavelength optical and near infrared images will map the distribution of late - type stellar populations .
the @xmath19 color maps will then highlight the regions of most recent star formation while @xmath20 will point to the somewhat older ( @xmath21 myr ) populations .
thus our slit positions generally cover the regions brightest in these color maps .
in general , the spectra from each slit position were subdivided into regions corresponding to some of the main features observed in the optical ground - based images , and one - dimensional spectra were extracted by summing pixels corresponding to the regions of interest .
the spectra of regions close to the qso nucleus were contaminated by scattered qso light .
the scattered light was removed by subtracting from each region a version of the quasar nuclear spectrum , scaled to match the broad - line flux . in the case of mrk1014
, none of our slit positions went through the qso , so we obtained a separate 200 s exposure of the nucleus .
spectra from those slit positions which actually went through the qso nucleus suffered from strong light scattering within the spectrograph , particularly in the spectral region around h@xmath22 .
therefore we were unable to obtain spectra of regions closer than @xmath23 from the qso nucleus for these slit positions .
wfpc2 images of the three objects were obtained from the _ hst _ data archive .
we used three 600 s wfc2 images of mrk1014 in the f675w filter , one 400 s and two 600 s pc1 images of iras07598 + 6508 in the f702w filter , and two 1100 s pc1 images of mrk231 in the f439w filter .
most cosmic rays were removed by subtracting a median image from each of the individual frames , then thresholding the difference at a 3@xmath24 level , setting points above this threshold to the median of the difference image .
pixels near the position of the peak of the qso were excluded from this process .
the corrected difference image was then added back to the median image , giving a corrected version of the original image .
the few cosmic rays within the relevant region that escaped this process were removed manually with the iraf task _ imedit_. in the case of mrk231 , where only two images were available , we first subtracted one from the other , and proceeded as above .
the procedure was repeated , interchanging the images .
all the corrected images for each object were then averaged .
we use @xcite isochrone synthesis models to fit the spectra of the host galaxies .
we will see in the following sections that the three objects which are the subject of this paper show strong evidence of having undergone some major tidal interaction . as we have described in our previous work ( cs97 , cs2000 ) , spectra of the host galaxies of such objects show features from both young ( _ e.g. , _ strong balmer lines ) and old ( _ e.g. , _ absorption ) stellar populations , and can usually be fitted satisfactorily by a two - component model .
this model includes an old underlying stellar population , presumably the stellar component present prior to interaction , and a younger instantaneous burst model , presumably produced as a result of the interaction .
a population with no age dispersion will be a reasonable approximation of the actual starburst as long as the period during which the star formation rate was greatly enhanced is short compared to the age of the population itself .
we have also noted previously ( scc98 ) that the age of the superposed starburst is remarkably robust with respect to the different assumptions about the nature of the older stellar component .
thus , we select a reasonable old underlying population ( with certain variations as described in each section below ) and assume that the same underlying population is present everywhere in the host galaxy . to this population we add instantaneous burst , @xcite initial mass function , solar metallicity models of various ages .
we then perform a @xmath25 fit to the data to determine the scaling of each component and the age of the most recent starburst .
the errors in the starburst ages that we quote are estimated by noting the youngest and oldest best fits for which @xmath26 changes by 15% with respect to the minimum value ( see cs2000 for details ) . in some cases , stellar absorption features ( most often the balmer lines )
were contaminated by emission coming from the extended narrow emission line region around the qso . in some cases , we subtracted a scaled synthetic spectrum of the recombination lines assuming case b. however , in calculating @xmath26 for the model fitting , we generally excluded those lines that suffered most from contamination .
all spectra are displayed as observed ( _ i.e. , _ without line subtraction ) , unless otherwise specified . as we discussed in cs2000 , because of our limited spatial resolution ( generally @xmath27 ) and projection along the line of sight , we are likely observing the integrated spectrum of several starbursts of different ages , and the age we determine will be somewhat older than the youngest starbursts . therefore the ages we report should be regarded as upper limits to the most recent episodes of star formation along the line of sight . in objects with recent starbursts ,
the effect of reddening by dust is an obvious concern .
however , studies of low - redshift agns and uligs at millimeter and submillimeter wavelengths indicate that dust is generally heavily concentrated within @xmath28 kpc of the nucleus @xcite .
in addition , have found ( scc98 ) that even in a case where the optical ir spectral index is strongly affected by dust , the stellar ages from spectral features in the rest - frame 32005200 region remain fairly robust .
this relative insensitivity to dust can be attributed to the fact that we are largely dealing with dust that is intermixed with the stars and that we preferentially observe regions with low extinction and , thus , low reddening . even in regions where dust along the line of sight is significant
, the reddening in our observed bandpass is likely to be largely compensated by blue light scattered into our line of sight .
mrk1014 ( @xmath29 ) is a luminous ( m@xmath30 ) , infrared loud ( _ e.g. , _ * ? ? ?
* ) radio - quiet qso which shows a luminous host galaxy .
the host galaxy has two large `` spiral - like arms '' ( @xcite ; see fig .
[ mrk1014mos ] and fig .
[ mrk1014hst ] ) .
its spectrum indicates a mixture of old and young stars @xcite .
recently , @xcite , in a spectroscopic survey of 26 rlqs , rqqs , and radio galaxies , observed the host galaxy of mrk1014 with the mayall 4 m telescope at kitt peak national observatory and with the 4.2 m william herschel telescope at la palma .
they modeled the spectra and determined an age of 12 gyr for the host galaxy of mrk1014 .
their approach is exactly opposite to ours ( see [ modeling ] ) : they fix the age of a possible young single starburst population to 0.1 gyr and let that of a second , older single starburst population ( what we would call the `` underlying population '' ) vary .
their reason for including the 0.1 gyr population is to account for `` the spectral shape of the blue light '' which they attribute to either a recent burst of star formation or contamination of the slit by scattered light from the qso nucleus .
their results will be discussed further in [ smrk1014spec ] .
the host galaxy of mrk1014 shows stellar absorption features with redshifts remarkably close to those of the qso broad and narrow emission lines ( @xmath31 , as measured from our spectrum ) . we have obtained and modeled spectra of different regions in the host galaxy of mrk1014 , and we shall refer to them according to their label in fig .
[ mrk1014slit ] .
we have chosen a 10 gyr old population with an exponentially declining star formation rate with an e - folding time of 3 gyr as an old underlying population .
this model fits reasonably well the spectrum of a galaxy 60 west - southwest of the qso at the same redshift and only slightly smaller than the `` bulge '' ( see below ) of mrk1014 .
( this galaxy was in our slit while obtaining the 200 s exposure of the qso ) .
the assumption that the host galaxy ( or parent galaxies ) of mrk1014 had a similar star formation history to this galaxy need not be accurate as the precise model we use as the pre - existing population makes little difference in the age determination of the starburst population ( scc98 ) . for comparison
, we have tried using a generic elliptical galaxy spectrum , and a model with a longer e - folding time ( _ i.e. , _ 5 gyr ) as underlying populations in the modeling of the spectra in mrk1014 .
we obtain the same starburst ages , though slightly different flux contributions from the old population , regardless of the model used .
figure [ mrk1014mos ] includes a @xmath32 color map of mrk1014 .
this image emphasizes those regions with a steeper blue continuum spectrum , peaking just redwards of the balmer limit .
these regions are concentrated mainly along the north edge of the tail ( regions @xmath33 and @xmath34 in fig .
[ mrk1014slit ] ) , in a clump on the east end of the tail ( @xmath35 ) , directly east of the nucleus ( not covered by our slits ) , and southwest of the qso nucleus ( @xmath36 ) .
spectroscopy of these regions confirms that they are indeed the youngest stellar populations we find in the host , with ages ranging from 180 myr in region @xmath36 to 290 myr in region @xmath37 .
figure [ mrk1014young ] shows the spectrum of region @xmath33 , with the best @xmath26 fit of the model to the data superposed , and the relative contributions of the 200 myr starburst and the old underlying population .
the error in these ages is typically @xmath38 myr .
figure [ mrk1014hst ] shows compact knots at the positions of regions @xmath33 through @xmath37 , and planetary camera _ hst _ images @xcite show a larger and very bright blue knot at the position of region @xmath36 .
even though there are knots at the positions of @xmath33 and @xmath34 , it is evident from colors and spectroscopy that there is recent star formation all along the north edge of the tail ( _ i.e. , _ between @xmath33 and @xmath34 ) .
other regions of the host galaxy appear redder in the @xmath32 color map in fig .
[ mrk1014mos ] .
these regions , sampled by @xmath39 and @xmath40 , appear to be dominated by an older , @xmath28 gyr population , and to have very little , if any , contribution from the old underlying population ( fig .
[ mrk1014older ] , top panel ) .
it is not entirely clear whether these are truly intermediate age populations , or if they are simply , as in the case of region @xmath41 , dominated by an old underlying population with a very small contribution from a younger population like those found in regions @xmath33 through @xmath36 .
we attempted to fit models with the latter characteristics to the observed spectra .
the resulting fits are reasonable , with minimum values of @xmath26 10% and 25% larger than that obtained for a dominant intermediate - age population for regions @xmath39 and @xmath40 respectively . the bottom panel of fig .
[ mrk1014older ] shows the spectrum of region @xmath39 with the best fit to the data of the sum of an old underlying population and a 250 myr instantaneous burst population .
while a reasonable fit , it does show significant discrepancies in the region around the 4000 break and the @xmath42 line .
the potential presence of this intermediate age population suggests that a better fit for the younger populations might be achieved by adding a third component that accounts for this intermediate age component .
however the flux of the young starburst population in those regions is so dominant ( typically contributing 80% of the total flux at rest - frame 5000 ) , that a third component makes a negligible difference to the fit .
region @xmath43 shows very strong emission , which almost certainly comes from gas ionized by the qso rather than from star - forming regions .
the equivalent widths of the emission lines here are greater than those of the emission lines elsewhere in the galaxy by at least a factor of 5 , and at least twice those of the qso nucleus for [ ] .
the emission line ratios indicate a power - law ionizing continuum @xcite .
this region is clearly seen as a large , discrete knot in fig .
2@xmath37 of @xcite .
the gas has an approaching velocity of @xmath44 km s@xmath5 with respect to the stellar absorptions in that region .
the underlying spectrum is similar to that of region @xmath41 .
region @xmath41 also shows emission lines with approaching velocities of @xmath45 km s@xmath5 with respect to the stellar features , but this region shows an additional , weaker emission component clearly visible in [ ] , [ ] , and h@xmath46 , blueshifted by @xmath47 km s@xmath5 with respect to the stronger component .
the stellar population here seems to be dominated by an old population with a small fraction of the flux coming from younger stars .
we have added spectra from two different slit positions at region @xmath48 to improve the signal to noise in this very faint region .
this is part of the long extension on the west side clearly seen in the high contrast image in fig .
[ mrk1014mos ] .
we find a continuum with a red sed and a clear 4000 break at a redshift close to , but slightly larger than that of the main body of the galaxy .
region @xmath49 , which is very bright in the @xmath50 image in fig .
[ mrk1014mos ] , would appear to be an extension of the east tail .
however , its spectrum shows that it is a background galaxy at @xmath51 .
likewise , region @xmath52 on the west tail is a galaxy at @xmath53 .
a third galaxy se of the qso ( @xmath54 , @xmath55 ) has a similar redshift as well ( @xmath56 ) , so there seems to be a group or cluster of galaxies at this redshift .
the 200 s exposure spectrum of the elongated object 92 west - southwest of the qso shows narrow emission lines at the same redshift as the qso superposed on a red continuum .
2@xmath37 of @xcite shows that there is strong emission just north of this object coming from extended gas ionized by the qso and not necessarily associated with the object .
the spectrum is too noisy to distinguish stellar features , so it is unclear whether this object is interacting with the system or if it is a chance projection .
the near edge - on galaxy @xmath57west of the qso , on the other hand , shows [ ] emission confined to the galaxy at @xmath58 in a university of hawaii 2.2 m telescope spectrum ( canalizo 2000 , unpublished ) , as suggested by @xcite from their [ ] imaging .
how do our results compare to those of @xcite ? as mentioned above , nolan et al .
use a fixed 0.1 gyr starburst population .
they determine that this population makes up 1.1% of the total luminous mass along the line of sight , and that the rest of the flux is well characterized by a 12 gyr instantaneous burst .
in contrast , we find several regions of recent ( @xmath59 gyr ) star formation where the star forming mass typically amounts to 12% , and sometimes up to 30% , of the total luminous mass along the line of sight .
one of our slit positions is very similar to one used by nolan et al.(see fig . 2 in @xcite ) , but our slit is narrower and slightly closer to the nucleus . as a way to compare , we added the flux along the slit as nolan et al . seem to have done , including the background galaxy @xmath49 , which is also in their slit .
even if we add up all the flux along this slit ( subtracting the qso light , which amounted to 5% of the total flux ) , we still find that the star forming regions make up 8% of the total luminous mass along the line of sight .
obviously , the difference in ages will lead to a smaller percentage for their choice of parameters .
so , we tried fixing our parameters to match theirs ( _ i.e. ,
_ 12 gyr + 0.1 gyr populations ) , and this yields a 5% by mass for the young population , but the fit is much inferior to the ones discussed in this section .
their slit position may have fortuitously missed the major star forming regions , thus leading to this smaller percentage . we have previously cautioned ( cs2000 ) that different slit positions can lead to different age determinations , and we emphasize the importance of carefully selecting slit positions if one wishes ( as we do ) to find the major starburst regions .
it is also far more difficult to obtain a reliable age for an older population in the presence of a contaminating younger population than the reverse .
images of mrk1014 show a very prominent tail extending to the northeast , reminiscent of the tidal tail of 3c48 ( cs2000 ) . like the tidal tail in 3c48 ,
this tail has a number of small clumps ( see _ hst _ image in fig .
[ mrk1014hst ] ) of star formation , which are commonly found in merging systems .
the bulk of the bright portion of the tail appears to be dominated by an intermediate age population ( @xmath141 gyr ) with a very small contribution , if any , from an old underlying population ; alternatively , it could be dominated by an older population , with some flux scattered from the bright star - forming regions ( @xmath33 through @xmath37 ) or fainter small regions distributed along the host
. however , the north edge of the tail appears as a very sharp feature in the short wavelength images , and contains stellar populations which are as young as those of the clumps .
as we find no redshift variations along this edge , it would appear that we are observing the tail nearly face - on , and that this is the leading edge where the material has been compressed , thus producing star formation .
this , again , is similar to the blue leading edge observed in 3c48 ( fig .
2@xmath36 in cs2000 ) , presumably sharper in mrk1014 because of the lower inclination angle .
both the ground - based and the _ hst _ images of mrk1014 ( figs .
[ mrk1014mos ] and [ mrk1014hst ] ) show a long , low - surface - brightness extension of the bright tail on the east of the nucleus and arching towards the south ( more evident after the removal of the background galaxy , @xmath49 ) as well a very extended faint secondary tail , rotationally symmetric to the bright ( primary ) tail .
each tail extends for as much as @xmath60 or @xmath14100 kpc ( note the inset in fig . [ mrk1014mos ] showing the well - known local interacting system m51/ngc5195 at the same scale ) .
spectra of the secondary tail west of the nucleus indicates that the tail is made up by older stars , and this is consistent with this tail being visible in our @xmath61 ( not shown ) and @xmath62 images .
no bright clumps of star formation are evident along the secondary tail ; this absence is not unusual as tidal dwarf formation appears not to be a ubiquitous process in mergers @xcite .
the _ hst _
image of the nucleus ( see inset in fig . [ mrk1014hst ] ) shows a small extension on the south side which follows the direction of the secondary tidal tail .
a similar extension is seen in _
hst nicmos _
images by @xcite .
there appears to be a `` bulge '' or enhanced brightness area elongated roughly along the axis connecting the beginning of both tails ( _ i.e. , _
se ) , with a half radius of @xmath63 kpc . assuming a projected velocity of 300 km s@xmath5 on the plane of the sky , the dynamical age for the tails is @xmath64 myr .
the tails are then older than every major post - starburst knot they contain , consistent with the idea that the latter were formed after the tails were first launched . at the same time , the tails are dynamically younger than the bulk of the stars that form them , though perhaps only slightly so if the stellar populations observed in @xmath39 and @xmath40 are truly @xmath28 gyr old , instead of being @xmath65 gyr with a small admixture of younger stars .
mrk1014 shows some striking similarities to 3c48 ( cs2000 ) : the morphology of the ( primary ) tidal tail , the clumps of star formation along the tail as well as its blue leading edge , the relation of the starburst ages to the dynamical age of the tails ( both ages @xmath66 myr younger in 3c48 ) , and the clumpy extended emission line region @xcite with high velocity ( @xmath67 km s@xmath5 ) components . as with 3c48 ,
the data strongly suggest that mrk1014 is the result of a merger of two galaxies of comparable size , both of which were disks in this case .
the starbursts in the main body of the host of mrk1014 , however , appear to be less intense and less widespread than those of 3c48 .
if the intermediate age ( 1 gyr ) population we see in regions @xmath39 and @xmath40 is real , it may be the relic of a starburst ignited at an initial passage of the two interacting galaxies .
indeed , we know of another system ( unj1025@xmath680040 ; * ? ? ?
* ) where the interacting galaxies have a difference in starburst ages of this order , possibly coincident with their orbital period .
if this were the case for mrk1014 , one might expect most of the star formation at the present to be very strongly concentrated towards the nucleus , as most of the gas would have been driven towards the center starting @xmath28 gyr ago .
the luminous , @xmath69 , radio - quiet qso iras07598 + 6508 was first detected by _ iras _ , identified as an agn candidate by @xcite , and spectroscopically identified as a qso by @xcite .
the optical spectrum of the qso is dominated by extremely strong emission @xcite , and the uv spectrum shows low- and high - ionization broad absorption lines ( bal ) extending to blueshifts of 5200 to 22000 km s@xmath70 @xcite .
figures [ ir0759mos ] and [ ir0759hst ] show , respectively , ground - based and _ hst _ images of iras07598 + 6508 ( see * ? ? ? * for a psf - subtracted version of the _ hst _ image ) .
these images show two clumpy regions @xmath71 south and southeast of the qso , bright in the psf subtracted @xmath16 image , but barely visible in @xmath61-band images ( not shown ) .
the great number of knots in these regions , presumably ob associations , already argues for recent star formation .
our slit positions cover these two regions , labeled @xmath33 and @xmath35 in fig .
[ ir0759slit ] , as well as some of the fainter emission surrounding the nucleus .
spectra of these regions are shown in fig .
[ ir0759spec ] .
regions @xmath33 and @xmath35 are fit by single starburst models of ages @xmath72 and @xmath73 myr respectively .
the light from these starbursts dominates the spectra as we do not find any significant contribution from an older component .
although single burst models fit the data better than two - component models , there are still some discrepancies in the fit .
the observed spectra show an excess with respect to the model in the region between 3900 and 4100 , apparently because the observed continuum is steeper in this region . to test whether the discrepancy could be an indication of scattered qso light , we subtracted qso spectra with several different scalings from the stellar spectra , but we were unable to obtain better fits .
in contrast to @xmath33 and @xmath35 , the spectra of the regions closer to the nucleus ( labeled @xmath37 and @xmath36 in fig .
[ ir0759slit ] ) show an older stellar population ( fig . [ ir0759spec ] , bottom panel ) .
the seds of these spectra are much redder and there is a clear , though not very prominent , 4000 break .
we detect this old population from @xmath74 to @xmath75 north of the qso and from @xmath74 to @xmath76 south of the qso .
we were unsuccessful in attempting to subtract the qso scattered light closer to the nucleus because of the strong light scattering along the slit in these regions .
therefore we are unable to determine whether the populations closest to the nucleus are as old as those of @xmath37 and @xmath36 or if there might be a younger starburst concentrated around the nucleus .
regions @xmath33 and @xmath35 have a slightly higher redshift than the fainter emission around the qso : @xmath77 and @xmath78 , compared to @xmath79 for the regions closer to the nucleus , but the difference is barely significant . measuring a precise redshift for the qso
is difficult because of the strong emission contaminating the broad emission lines , and the absence of narrow emission lines .
values in the literature include @xmath80 as measured from co @xcite and @xmath81 as measured from broad balmer emission lines @xcite ; both are within 180 km s@xmath5 ( and within the errors ) of the values we measure for the host galaxy . iras07598
+ 6508 has some limited extended narrow emission as seen in the [ ] image in fig .
[ ir0759mos ] .
the ionized gas seems to be correlated with the stellar component : the velocity difference between emission and absorption lines in regions where both are present is never larger than 50 km @xmath82 ( notice the profile of h@xmath46 in fig .
[ ir0759spec ] ; emission is visible slightly redshifted with respect to the absorption ) .
region @xmath34 shows the strongest emission and appears brightest in the [ ] image .
the emission line spectrum has an underlying blue continuum with stellar features .
the 2-dimensional spectrum shows that this region is broken up into two discrete clumps a little over 1 each , having some small scale velocity structure .
the line flux ratios indicate that the south clump has lower ionization , and it may be an region rather than gas ionized by the qso .
a second , fainter component with a velocity gradient from 0 to @xmath83 km @xmath82 apparently originating from the north clump extends 1 towards the south .
about 11 north of the qso , we find another emission region ( labeled @xmath40 ) which is visible in the [ ] image , but not in any of the broad band images .
no stellar continuum is evident from the spectrum , either .
this region is also broken into two discrete , somewhat larger clumps , but in this case the clump closer to the nucleus has lower ionization .
both clumps are at a lower redshift than those of region @xmath34 , _
i.e. , _
@xmath84 _ vs. _ @xmath85 in @xmath34 . our deep @xmath50 image ( fig .
[ ir0759mos ] ) shows a tidal tail extending from north to the east and arching towards the south of the nucleus for @xmath86 or @xmath87 kpc .
the dynamical age of this feature ( assuming a projected velocity of 300 km @xmath82 ) is @xmath88 myr , again older than the starbursts ages found for this object .
we have a very faint spectrum of this tail at region @xmath39 .
the sed of the spectrum is similar to that of @xmath37 and @xmath36 , but the spectrum is too noisy to show stellar features . an [ ]
@xmath89 emission line is confined to this region ( as seen in the 2-d spectrum ) at a redshift of @xmath90 , which is probably close to that of the stellar component as in other regions .
this tidal tail is strongly suggestive of a merger .
the fact that only one tail is evident may indicate that we are seeing the result of a merger of a spiral with an elliptical galaxy .
this configuration , however , could also result from the merger of two spiral galaxies , one of which is counter rotating to the relative orbit . in mergers with such geometry ,
the gas and stars in the counter - rotating disk are only slightly perturbed and are not pulled into tidal bridges and tails @xcite .
it is also possible that a second tail may not be evident because of projection effects .
regions @xmath33 and @xmath35 may be part of the host galaxy with enhanced surface brightness due to the recent star formation , or they may be remnants of companion galaxies which have strongly interacted with the host galaxy . in either case
it is clear that these regions are tidally disturbed and they will likely be completely mixed with the host within a few crossing times .
the two galaxies @xmath91 south of the qso are likely to be companion galaxies since they are bright in our [ ] image ( fig .
[ ir0759mos ] ) , but we do not have a spectrum to confirm this .
the colors for the se galaxy indicate that it may have recent star formation , if at the same redshift as iras07598 + 6508 .
we find two additional objects with the same redshift as iras07598 + 6508 that happened to fall in our slits : a faint emission line object @xmath92 south - southwest of the qso , and a bright galaxy with an absorption and emission line spectrum @xmath93 southwest of the qso .
mrk231 ( @xmath94 ) , often classified as a seyfert 1 galaxy , is slightly below the luminosity cutoff for qsos defined by @xcite .
however , the nucleus is heavily reddened , with an estimated @xmath95 of foreground extinction ( * ? ? ?
* ; * ? ? ?
* ; see also * ? ? ?
* ) , so it would be well above this threshold if it were unobscured .
the central plateau of the host galaxy is off center with respect to the nucleus , and the presence of tails to the north and south , as well as a low - surface - brightness extension to the east clearly indicate a recent merger ( fig .
[ mrk231mos ] ) .
closer to the nucleus , _ hst _ imaging shows a large number of stellar associations indicating recent star formation ( @xcite ; see fig .
[ mrk231hst ] ) . as in the case of iras07598 + 651 ,
the spectrum of the qso is dominated by strong emission @xcite and shows a peculiar low - ionization bal system with velocities up to @xmath96 km s@xmath5 @xcite .
previous spectroscopy of the host galaxy indicates the presence of a young stellar population @xcite .
we have obtained and modeled spectra of several regions in the host galaxy as labeled in fig .
[ mrk231slit ] . in regions @xmath49 and @xmath43 , we find no evidence for a significant young starburst population . a 10 gyr - old population with an exponentially decreasing sfr and an e - folding time of 5 gyr fits this spectrum quite well , as shown in the bottom panel of fig .
[ mrk231spec ] .
therefore we use this model as an underlying population to fit the rest of the spectra in the host galaxy . unfortunately , because of the poor sensitivity of lris shortwards of 4000 , and the fact that mrk231 has a lower redshift ( @xmath97 ) than the other two objects , we were unable to obtain the near - uv portion of the spectrum , which is the most helpful region in determining the ages of post - starburst populations .
we are left then with the region on the long side of 3800 , where more than one combination of models can give similar fits to the spectra .
indeed , we found degenerate fits in some regions of the host galaxy , particularly in those regions with the youngest populations and those where the absorption lines were heavily contaminated by emission so that we could not use balmer lines to discriminate between models .
in such regions we obtained two @xmath26 minima , generally with one corresponding to a model with a small contribution ( _ i.e. , _ small percentage of total mass along the line of sight ) from a very young starburst , and the other to a model with a large contribution from a somewhat older starburst .
we illustrate the problem in fig .
[ degenerate ] , where we have plotted two models resulting from a 4 myr and a 42 myr old populations contributing , respectively , 1% and 12% of the total luminous mass .
the models are nearly identical in the spectral region redwards of 3800 , and both are good fits to the data ( see inset in fig . [ degenerate ] ) with @xmath26 values for these models differing only by 10% .
the models , however , diverge quickly at shorter wavelengths . in order to discriminate between the two `` best fits '' to the data
, we have measured photometry of the different regions ( @xmath33 through @xmath49 as indicated in fig .
[ mrk231slit ] ) from our ground - based optical images .
we have been careful to measure fluxes only within the areas limited by the slits , at similar resolutions , and subtracting scattered light from the qso where necessary .
the solid circles in fig .
[ degenerate ] indicate the photometry of region @xmath33 . clearly , the older model ( red line ) is a better fit to the data .
it is important to note , however , that if there is considerable reddening by dust along the lines of sight to these regions in the host galaxy , the flux values of @xmath16 will be depressed ; therefore the photometric points can only help us to determine upper limits to the ages of the stellar populations . for consistency ,
the age errors quoted in this section are as defined in
[ modeling ] , and do not take into account additional constraints placed by the photometry .
region @xmath33 corresponds to the west side of the arc - shaped structure ( the horseshoe ; * ? ? ?
* ) , @xmath63 south of the qso nucleus .
the _ hst _
f702w image in fig .
[ mrk231hst ] shows that region @xmath33 is formed by multiple knots , much like those of regions @xmath33 and @xmath35 in iras07598 + 6508 .
however this structure , unlike those of iras07598 + 6508 , has a remarkably similar morphology from the near uv to the near ir as shown in the top right panel of fig .
[ mrk231mos ] .
as discussed above , we determine an age of 42 ( + 22 , -17 ) myr .
this age is younger than the 225 myr estimated from @xmath98 colors of this region by @xcite . while several authors have noted the blue color of the `` horseshoe '' ( _ e.g. , _ *
* ; * ? ? ?
* ) , we find a region that is relatively much brighter at shorter wavelengths 16south of the qso , labeled @xmath35 in fig .
[ mrk231slit ] .
this region appears as a bright blue extended blob dominating the @xmath19 color map in fig .
[ mrk231mos ] .
the spectrum has a very steep blue continuum , and very strong emission lines at the same redshift as the absorption lines , likely from an hii region @xcite .
the top panel of fig .
[ mrk231spec ] shows the spectrum of region @xmath35 with the balmer emission lines subtracted as described in
[ modeling ] .
we find an age for this region of 5 myr .
as such a young starburst age is comparable to the expected duration of a typical individual starburst , and we are likely observing a collection of several such starbursts , it is virtually impossible to distinguish between continuous star formation and instantaneous bursts .
the age we find is , therefore , simply indicative of ongoing star formation . as in region @xmath33
, we find degeneracy in the modeling of this spectrum , with a second @xmath26 minimum at 100 myr .
the near - uv continuum of the older model , however , falls @xmath99 below the @xmath16 photometric point .
neither the deep optical ground - based images nor the _ hst _ image show evidence for stellar associations in region @xmath35 .
this is somewhat surprising since , although the starburst population contributes only 1% to the luminous mass along the line of sight , its flux amounts to 42% of the total flux at 5000 (rest wavelength ) , and even more at shorter wavelengths .
regions @xmath34 and @xmath36 appear as single clumps in the images and have ages of 140 ( + 80 , @xmath6870 ) myr and 180 ( + 60 , @xmath6880 ) myr respectively .
the starburst populations dominate the spectra in these regions , contributing @xmath100% of the total flux at rest frame 5000 and up to 36% of the total luminous mass along the line of sight . regions @xmath37 , @xmath39 ( fig . [ mrk231spec ] ) , @xmath40 , and @xmath41 all show very similar spectra , as expected from the color maps ( fig .
[ mrk231mos ] ) , with ages between 300 and 360 myr .
the starbursts in these regions typically contribute only @xmath1450% to the total flux and @xmath101% of the total luminous mass , except for region @xmath37 , for which these values are very similar to those of regions @xmath34 and @xmath36 .
we detect very weak continuum from the north tail ( region @xmath52 ) at the @xmath102 level .
the continuum is slightly blue , and there is a hint of @xmath34 at the redshift of the host galaxy .
this region corresponds to the blue region in the @xmath103 color map , just before the condensation at the end of the north tail .
it is possible that this region contains a knot of star formation like those found in the tail of mrk1014 .
@xcite reports the presence of an extended bal in the west side of the host galaxy . the only place where we see evidence for this feature is in region @xmath48 .
the spectrum of @xmath48 , uncorrected for qso scattered light , shows a wide absorption feature ( @xmath104 km s@xmath5 fwhm compared to @xmath105 km s@xmath5 for the qso bal ) as well as narrow absorption and emission ( see below ) at the redshift of the stellar component . after correcting for qso contamination
, the feature appears narrower , but still at @xmath1066400 km s@xmath5 with respect to the emission line .
the hypothetical bal feature is , however , much weaker than that of the qso , and it could be an artifact of an imperfect subtraction of the qso light . while the evidence for extended bal is not strong in our data ,
@xcite find an `` excess flux '' extending to the blue side of [ ] @xmath1076548 , which they interpret as a broad blueshifted component of h@xmath22 or h@xmath22 + [ ] indicative of an outflow with velocities up to 1500 km s@xmath5 .
we observe this excess most clearly in regions @xmath41 and @xmath40 extending to even larger velocities ( @xmath108 km s@xmath5 ) .
@xcite find some level of excess light in the blue wing of the h@xmath22 line in every low - ionization bal qso in their sample and suggest that this excess flux could be from the bal material itself . the narrow emission line shows a peculiar behavior .
we observe narrow emission in regions @xmath41 and @xmath48 on the red side of the absorption line , forming what looks like a p - cygni profile .
the only other emission lines evident in region @xmath41 are [ ] @xmath1076583 , and weak [ ] @xmath1096717,6731 , but these lines are slightly _ blueshifted _ with respect to the absorption lines . in region
@xmath40 the emission line disappears , but the absorption line becomes very weak as well , so it is possible that the emission line is blueshifted into the absorption line ; [ ] , however , is at a higher redshift than in region @xmath41 .
thus , the very low ionization gas appears to be decoupled from the moderately low ionization gas in these regions .
@xcite find a `` green '' ( _ i.e. , _ visible only in their @xmath110 image ) , jet - like feature extending from near the nucleus to the northeast .
they suggest it `` may be line emission in [ ] , perhaps tracing ionizing radiation originating in the nucleus '' .
our slit goes through the western side of where this feature would be ( see fig . 2 in
* ) , and region @xmath48 should sample the brightest part of the ridge leading to the bright knot .
however , we find only weak [ ] in the spectrum of this region , certainly much weaker than in other regions we sample . therefore it is unlikely that this `` green '' feature , if real , is due to ionized gas .
images of mrk231 show a greatly disturbed host galaxy both in large and small scales ( fig . [ mrk231mos ] ) .
at large scales , two symmetric tails extend on the east side of the nucleus for @xmath111 or @xmath112 kpc each , and there is some very extended low surface brightness material east of the north tail .
the nucleus is off - center on a bright `` plateau '' @xcite 20 kpc across that shows complex morphology , including linear jet - like structures ( region @xmath40 ) , and curved tail - like structures ( region @xmath37 ) .
closer to the nucleus ( fig .
[ mrk231hst ] ) we find the `` horseshoe '' 3 kpc to the south with numerous clumps of star formation , and arm - like features spiraling around the nucleus which connect to the outer tidal tails @xcite this morphology is indicative of a merger between two disk galaxies of similar mass .
while we are unable to obtain stellar spectra within the central kpc ( @xmath113 ) of mrk231 , @xcite have found from uv spectropolarimetry of the nuclear region that high polarization in the optical falls off quite rapidly shortward of @xmath114 ; this effect is most easily interpreted as a dilution of the polarized component by o and b stars from an ongoing starburst @xcite .
co and radio observations also show evidence for a centrally concentrated starburst @xcite .
it is possible that the knots along the arm - like features around the nucleus have ages similar to those of region @xmath33 , if not younger .
we find a relatively wide range of starburst ages around the host galaxy . if , once again , we assume a projected velocity of 300 km s@xmath5 , the tidal tails have a dynamical age of 110 myr .
mrk231 is unusual in that there seems to be a significant post - starburst population somewhat older ( 300360 myr ) than the tails , which may indicate that wide spread star formation was ignited prior to the stages of final merger .
numerical simulations of mergers @xcite predict such a scenario when the merging galaxies lack a significant bulge to stabilize them against an early dissipation .
in cs2000 we noted that the apparent correlation between bulge mass and black - hole mass @xcite suggests a possible correlation between qso luminosity and the delay in star formation activity .
thus , the early starburst activity in mrk231 may indicate that the host galaxy did not have a substantial bulge , which would in turn imply a less massive black - hole that may result into a less luminous agn .
although we do in fact observe a relatively less luminous agn in mrk231 , these connections are highly speculative , and they are based on relations @xcite and models @xcite that still have many uncertainties .
mrk1014 , iras07598 + 6508 , and mrk231 show many similarities .
we already knew that these three objects , while being optically or ir selected qsos , were also part of the ulig family . in this study
we have found additional properties which they share and which may be related to their intermediate position in the fir diagram .
while mrk1014 and mrk231 have long been known to have tails , we have shown that iras 07598 + 6508 also has at least one tidal tail .
we have found evidence that all three objects have undergone a strong interaction and are now in the final stages of mergers .
the morphology of the hosts ( highly perturbed galaxies with tidal tails and destroyed disks ) as well as the extent of the starbursts indicate that these are all major mergers between galaxies of comparable mass rather than accretion events of low mass dwarf companions @xcite .
all three objects show spectra typical of e+a galaxies @xcite , that is , spectra characterized by the simultaneous presence of strong balmer absorption lines indicative of a young stellar population and features from an older population such as , and the absence of strong emission lines typical of star forming galaxies ( we reiterate that the emission lines seen in our spectra generally come from extended gas ionized by the qso rather than from star forming regions , with a few exceptions ) .
e+a galaxies are frequently linked to galaxy - galaxy mergers and interactions ( _ e.g. , _ * ? ? ?
* ) . in the case of objects with e+a spectra undergoing tidal interactions ,
the young superposed population is clearly related to the interaction .
the three objects discussed in this paper , along with 3c48 ( cs2000 ) and pg1700 + 518 ( cs97 ) , all have interaction - induced starbursts .
in addition to the similarities noted in the host galaxies of the three objects , there are also some common characteristics in the spectra of the active nuclei themselves .
all three objects show strong emission , with two of these objects ( iras07598 + 6508 and mrk231 ) being `` extreme '' emitters ( _ i.e. , _ showing ratios of @xmath115(feii @xmath116)/@xmath115(h@xmath46 ) @xmath117 2 ; @xcite ) .
the latter two objects , like pg1700 + 518 ( cs97 ; scc98 ) , are also low - ionization bal qsos and have weak or absent narrow emission lines .
how do the ages of these starbursts relate to the merger / interaction stage ?
the dynamical ages for tidal features in these objects are uncertain because of projection effects .
however , our rough estimates indicate that in two cases , mrk1014 and iras07598 + 6508 , the peak of the starburst occurred after the tidal tails were launched .
this requires some mechanism to stabilize the gas contents of the galaxies against bursting in star formation until the later stages of the merger ; numerical simulations indicate that a significant bulge in the host can provide this mechanism @xcite .
mrk231 , on the other hand , shows starburst ages which indicate that much of the star formation activity may have started before the last stages of the merger ; hence , one or both of the merging galaxies may have lacked a significant bulge . whenever we have found a significant range of starburst ages in the host galaxies of qsos , we have also found evidence that the youngest major starburst regions are preferentially concentrated towards the center of the galaxy . in 3c48 ( cs2000 )
we found the clearest example of this , with starburst ages becoming progressively younger and more dominating as we approached the galaxy / qso nucleus .
mrk231 shows older starbursts in the `` plateau '' extending @xmath118 kpc around the nucleus , with some of the youngest populations only 3 kpc from the nucleus , and possibly even younger populations in the central kpc of the galaxy .
mrk1014 shows stronger relatively recent star formation activity along the tail than any of the other objects . however , the starburst region we observe within 2 kpc of the nucleus is far more massive , luminous , and larger @xcite than the starburst regions along the tail , and has the youngest age found in the host .
the interaction histories of the five objects discussed so far clearly favor a strong connection between interactions and vigorous bursts of star formation .
since the gas flows towards the inner regions can not only trigger star formation but also serve as fuel to the qso , one might expect the age of the qso activity to be closely related to the age of the initial starbursts in the central regions of the galaxy .
however , observations of central starbursts are in every case hampered by the presence of the qso ; even if this region could be observed , the spectrum would likely be dominated by continuing recent starburst activity and not by a starburst that was coincident with the onset of the qso activity .
furthermore , both starburst and qso activity may be episodic .
all of this is to say that there is an unavoidable intrinsic uncertainty in using starburst ages to place qsos in an evolutionary sequence .
we defer the detailed discussion of a possible age sequence to a subsequent paper where we will present the results of the four remaining objects in our sample ( _ i.e. , _
iras00275@xmath682859 , iras04505@xmath682958 , izw1 , and pg1543 + 489 ) .
we emphasize , though , that many of our general conclusions from the objects discussed here , as well as from our observations of 3c48 ( cs2000 ) , have an interest that is quite independent of any attempt to use some sort of starburst age as a proxy for a qso age .
( 1 ) the confirmation that these are all starburst or post - starburst objects and that they all show obvious tidal tails validates their close connection with other uligs , virtually all of which are mergers or strongly interacting pairs .
while there has long been strong _ circumstantial _ evidence that a large fraction of the qso population has resulted from triggering of the qso activity by interactions and mergers , we now have much more direct evidence for this mechanism for at least one subclass of qsos .
( 2 ) both the spatial distribution and the time history of star formation in a qso host galaxy give clues to nature of the galaxies that have participated in the merger .
while more sophisticated models of star formation during interactions will be necessary to exploit these data fully , we already have some hints in terms of the enhanced star formation along the leading edge of the tails in 3c48 and mrk1014 , and in the relative ages of the tail structures and the star - forming regions contained within them .
( 3 ) the youth of the stellar populations in these objects reinforces previous suggestions connecting strong emission and low - ionization bal features with the relatively recent triggering of qso activity .
we will discuss these connections in detail in the paper presenting the observations of the remaining four objects in our sample .
we thank gerbs bauer , scott dahm , and susan ridgway for assisting in some of the observations , and bill vacca for helpful discussions about imfs .
we also thank the referee , dean hines , for his very prompt review of the paper and his suggestions , which helped us improve both its content and its presentation .
this paper was partly written while both authors were visitors at the research school of astronomy and astrophysics of the australian national university , and we thank both the director , jeremy mould , and the staff there for their hospitality .
this research has made use of the nasa / ipac extragalactic database ( ned ) which is operated by the jet propulsion laboratory , california institute of technology , under contract with the national aeronautics and space administration .
this research was partially supported by nsf under grant ast95 - 29078 .
andreani , p. , franceschini , a. , & granato , g. 1999 , , 306 , 161 boksenberg , a. , carswell , r. f. , allen , d. a. , fosbury , r. a. e. , penston , m. v. , & sargent , w. l. w. 1977 , , 178 , 451 boroson , t. a. , & meyers , k. a. 1992 , , 397 , 442 boyce , p. j. et al .
1996 , , 473 , 760 bruzual a. , g. & charlot , s. 1996 , unpublished bryant , p. & scoville , n. z. 1996 , , 457 , 678 canalizo , g. , stockton , a. , brotherton , m. s. , & van breugel , w. 2000 , aj , 119 , 59 canalizo , g. , & stockton , a. 2000 , , 528 , 201 [ cs2000 ] canalizo , g. , & stockton , a. 1997 , , 480 , l5 [ cs97 ] clements , d. l. et al .
1996 , , 279 , 459 de grijp , m. h. k. , miley , g. k. , & lub , j. 1987 , , 1987 , 70 , 95 downes , d. , & solomon , p. m. 1998 , , 507 , 615 dressler , a. & gunn , j. e. 1983 , , 270 , 7 foster , k. , rich , r. m. , & mccarthy , j. k. 1995 , , 450 , 74 goodrich , r. w. & miller , j. s. 1994 , , 434 , 82 gunn , j. e. 1979 in _ active galactic nuclei _
, ed . c. hazard and s. mitton ( new york : cambridge university press ) , p. 213
hamilton , d. & keel , w. c. 1987 , , 321 , 211 heckman , t. m. , bothun , g. d. , balick , b. , & smith , e. p. 1984
, , 89 , 958 hibbard , j. e. & yun , m. s. 1999 , , 118 , 162 hines , d. c. , low , f. j. , thompson , r. j. , weymann , r. j. , & storrie - lombardi , l. 1999 , , 512 , 140 hines , d. c. & wills , b. j. 1995 , , 448 , 69 hughes , d. h. , kukula , m. j. , dunlop , j. s. , boroson , t. 2000 , , in press [ astro - ph/0002021 ] hutchings , j. b. , & crampton , d. 1990 , , 99 , 37 hutchings , j. b. , & neff , s. g. 1987 , , 92 , 14 hutchings , j. b. , & neff , s. g. 1992 , , 104 , 1 kodaira , k. , iye , m. , & nishimura , s. 1979 , , 31 , 451 lawrence , a. et al . 1988 , , 235 , 261 lpari , s. 1994 , 436 , 102 lpari , s. , colina l. , & macchetto f. , 1994 , 427 , 174 low , f. j. , huchra , j. p. , kleinmann , s. g. , cutri , r. m. 1988 , , 327 , l41 mackenty , j. w. , & stockton , a. n. 1984 , , 283 , 64 mcleod , k. k. , & rieke , g. h. 1994 , , 431 , 137 magorrian , j. , et al . 1998 , , 115 , 2285 massey p. , strobel k. , barnes , j. v. , & anderson , e. 1988 , , 328 , 315 mihos , j. c. , & hernquist l. 1996 , , 464 , 641 mihos , j. c. , & hernquist l. 1994 , , 425 , l13 neugebauer , g. , soifer , b. t. , & miley , g. k. 1985 , , 295 , l27 neugebauer , g. , miley , g. k. , soifer , b. t. , & clegg , p. e. 1986 , , 308 , 815 nolan , l. a. , dunlop , j. s. , kukula , m. j. , hughes , d. h. , boroson , t. , jimenez , r. 2000 , , in press [ astro - ph/0004325 ] oke , j. b. , et al . 1995 , , 107 , 375 sanders , d. b. , soifer , b. t. , elias , j. h. , madore , b. f. , matthews , k. , neugebauer , g. , & scoville , n. z. 1988 , , 325 , 74 sanders , d. b. , scoville , n. z. , & soifer , b. t. 1988 , , 335 , 1 scalo , j. m. 1986 , fund .
cosmic phys . , 11 , 1 scoville , n. z. , evans , a. s. , thompson , r. , rieke , m. , hines , d. c. , low , f. j. , dinshaw , n. , surace , j. a. , armus , l. 2000 , , 119 , 991 schlegel , d. j. , finkbeiner d. p. , & davis , m. 1998 , , 500 , 525 schmidt , r. , & green , r. f. 1983 , , 269 , 352 smith , p. s. , schmidt , g. d. , allen , r. g. , & angel , j. r. p. 1995
, , 444 , 146 solomon p. m. , downes , d. , radford , s. j. e. , & barrett , j. w. 1997 , , 478 , 144 stockton , a. 1999 , in galaxy interactions at low and high redshift , iau symp . 186 , eds .
d. sanders & j. barnes ( san francisco : astron .
soc . pac . ) , p. 311
stockton , a. 1982 , , 257 , 33 stockton , a. , canalizo , g. , & close , l. 1998 , , 500 , l121 [ scc98 ] stockton , a. , & mackenty 1987 , , 316 , 584 surace , j. a. , sanders , d. b. , vacca , w. d. , veilleux , s. , & mazzarella , j. m. 1998 , , 492 , 116 surace , j. a. & sanders , d. b. 2000 , , in press [ astro - ph/0005122 ] taniguchi , t. , kawara , k. , nishida , m. , tamura , s. , nishida , m. t. 1988 , , 95 , 1378 taylor , g. b. , silver , c. s. , ulvestad , j. s. , & carilli , c. l. 1999 , , 519 , 185 toomre , a. , & toomre , j. 1972 , , 178 , 623 veilleux , s. & osterbrock , d. e. 1987 , , 63 , 295 zabludoff , a. i. et al . 1996 , , 466 , 104 lclccc mrk1014 & 103.0 & 4.0 n & 2.44 & 3600 & 97 sep 13 + mrk1014 & 87.1 & 6.9
n & 2.44 & 3600 & 98 sep 01 + mrk1014 & 148.6 & 2.2 w & 2.44 & 2400 & 98 sep 01 + iras07598 + 6508 & 243.6 & 6.9 s & 1.28 & 3600 & 96 oct 13 + iras07598 + 6508 & 190.0 & 0.0 & 1.28 & 3600 & 96 oct 14 + mrk231 & 6.0 & 0.0 & 2.44 & 1800 & 97 jun 12 + mrk231 & 83.0 & 16.0 s & 2.44 & 3600 & 98 mar 21 + mrk231 & 30.0 & 7.0 w & 2.44 & 2400 & 98 mar 21 + llcccc mrk1014 & orbit ccd & @xmath16 & 8@xmath151200 & 0.138 & 97 nov 01 + mrk1014 & orbit ccd & @xmath119 & 5@xmath15300 & 0.138 & 97 oct 31 + mrk1014 & tek1024 ccd & @xmath50 & 5@xmath15300 & 0.222 & 91 nov 13 + mrk1014 & quirc & @xmath61 & 29@xmath15300 & 0.061 & 99 oct 29 + mrk1014 & nicmos-3 & @xmath62 & 39@xmath1570 & 0.374 & 90 oct 28 + iras07598 + 6508&loral ccd&@xmath16 & 8@xmath151200 & 0.138 & 97 mar 09 + iras07598 + 6508&tek1024 ccd&@xmath50 & 15@xmath15300 & 0.222 & 92 mar 01 + iras07598 + 6508&tek1024 ccd&[oiii ] & 7@xmath151560 & 0.222 & 92 feb 27 + iras07598 + 6508 & quirc & @xmath61 & 22@xmath15120 & 0.061 & 99 oct 30 + mrk231 & loral ccd & @xmath16 & 5@xmath151200 & 0.138 & 97 mar 08 + mrk231 & loral ccd & @xmath120 & 18@xmath15200 & 0.138 & 97 mar 09 + mrk231 & loral ccd & @xmath121 & 6@xmath15300 & 0.138 & 97 mar 08 + mrk231 & quirc & @xmath61 & 69@xmath1520 & 0.061 & 99 apr 04 + | we present deep spectroscopic and imaging data of the host galaxies of mrk1014 , iras07598 + 6508 , and mrk231 .
these objects form part of both the qso and the ultraluminous infrared galaxy ( ulig ) families , and may represent a transition stage in an evolutionary scenario .
our imaging shows that all three objects have highly perturbed hosts with tidal tails and destroyed disks , and appear to be in the final stages of major mergers .
the host galaxies of the three objects have spectra typical of e+a galaxies , showing simultaneously features from an old and a young stellar component .
we model spectra from different regions of the host galaxies using @xcite spectral synthesis models using two component models including an old underlying population and recent superposed starbursts .
mrk1014 has intense star formation concentrated in a large knot @xmath0 kpc from the nucleus , along the leading edge of the tidal tail , and in several knots scattered around the host .
the starburst ages in these regions range from 180 to 290 myr .
iras 07598 + 6508 has multiple knots of star formation concentrated in two regions within 16 kpc of the qso nucleus , with ages ranging from 30 to 70 myr ; the host galaxy shows an older population in other regions .
mrk231 shows a wider range of starburst ages , ranging from 42 myr in the arc 3 kpc south of the nucleus , to @xmath1 myr spread on a `` plateau '' @xmath2 kpc across around the nucleus , as well as a uv bright region 12 kpc south of the nucleus , which is apparently a region of currently active star formation .
our results indicate a strong connection between interactions and vigorous bursts of star formation in these objects .
we propose that the starburst ages found are indicative of young ages for the qso activity .
the young starburst ages found are also consistent with the intermediate position of these objects in the far infrared color - color diagram . |
planets discovered using the radial velocity ( rv ) method have dominated the total exoplanet count until recently , when the transit method has made increasing contributions .
the long time baseline of rv surveys has allowed the detection more diverse orbital geometries than achievable by ground - based transit surveys .
the kepler mission , however , with its multi - year baseline , can begin to probe into parameter space previously reserved for rv studies . at longer periods ,
orbits tend to diverge significantly from the circular case beyond a semi - major axis of @xmath0 au @xcite , although there may be small observational biases that skew this distribution @xcite .
this insight has led to numerous attempts to account for eccentricity in the context of planet formation and orbital stability @xcite and the influence of tidal circularization @xcite .
it has been shown how eccentricity distribution effects transit probabilities @xcite and projected yields of transit surveys @xcite .
this influence is minor for the ground - based surveys since they are primarily sensitive to giant planets in short - period orbits .
however , the kepler mission is expected to be impacted by this distribution since it probes out to much longer periods with a much reduced disadvantage of a window function that affects observations from the ground @xcite .
a comparison of the kepler results in the context of eccentricity and transit durations with the rv distribution has been suggested by @xcite and @xcite and carried out by @xcite , but initial planet candidate releases by the kepler project do not provide enough period sensitivity @xcite .
the most recent release of kepler planet candidates by @xcite increases the total number of candidates to more than 2,300 and the time baseline probed to beyond 560 days .
this has several implications for studies of eccentricity distributions .
the kepler mission is sensitive to planets significantly smaller than those accessible by current rv experiments and thus allows a more in - depth study of the dependence of eccentricity on the planet mass / size and multiplicity .
if the eccentricity distributions of kepler and rv planets were found to be substantially different then this may reveal a selection effect in the way kepler candidates are selected which is biased against eccentric orbits .
a direct comparison of the two distributions , provided they are consistent for the planet mass / size region where their sensitivities overlap , will allow a more exhaustive investigation of orbital eccentricity to be undertaken . here
we present a study of the eccentricity distribution of planets discovered with the rv method and the complete list of kepler planet candidates .
we calculate expected transit durations for circular orbits and compare them with either calculated or measured eccentric transit durations ( [ eqns ] ) .
our results show that the measured transit durations from rv data ( [ rv ] ) and the kepler candidates ( [ kepler ] ) are consistent with having the same distribution .
we estimate the impact parameter distribution for the kepler candidates and show that their mean eccentricity decreases with decreasing planet size ( [ correlation ] ) , which supports the hypothesis that smaller planets tend to be found in multiple systems in near - circular orbits .
we discuss additional astrophysical aspects in [ discussion ] and conclude in [ conclusion ] .
a concise description of exoplanetary transit modeling and associated parameters is presented elsewhere @xcite . here
we concentrate on the relevant details to our analysis : transit duration and eccentricity . the transit duration for a circular orbit
is given by @xmath1 where @xmath2 is the orbital period , @xmath3 is the semi - major axis , @xmath4 is the orbital inclination , and @xmath5 and @xmath6 are the stellar and planetary radii respectively .
the impact parameter of a transit is given by @xmath7 and is defined as the projected separation of the planet and star centers at the point of mid - transit . [ cols="^,^ " , ] finally , we investigate a sample of the outliers with particularly large deviations from the circular model ( @xmath8 hours ) .
these candidates are shown in table [ ecctab ] .
since the kepler data frequently do not contain any secondary eclipse , @xmath9 and @xmath10 are unknown .
we calculate transit duration @xmath11 as a function of @xmath9 and @xmath10 via equation [ scaling ] .
we then produce a grid of @xmath12 for all values of @xmath9 and @xmath10 .
locations where the grid values are approximately equal to 1 are possible solutions for which the measured transit duration in the kepler candidate catalog is consistent with certain values of @xmath9 and @xmath10 .
an example of this is shown in figure [ evsw ] where we present results of the above calculations as an intensity map for koi 1230.01 . in order to be compatible with the kepler measured duration
, the eccentricity of the planet must be at least 0.34 .
this process is repeated for each of the candidates in table [ ecctab ] in which we report the minimum required eccentricity @xmath13 for each candidate .
it is worth noting , however , that these minimum eccentricities are not singular values but rather distributions , as can be seen by the gray - scale in figure [ evsw ] .
the uncertainties depend highly upon the various random errors in the measured values of the kepler candidates catalogue , including @xmath4 .
for example , the stellar radius of koi 2481.01 would need to be @xmath14% of the catalogue value in order for it to be in a circular orbit and the duration discrepancy to be reduced to zero .
using @xmath15 by @xmath16 using @xmath11 .
thus , a value of 1 ( peak intensity ) corresponds to the best solution ( [ kepler]).,width=309 ] further of interest in table [ ecctab ] are the relatively short - period planets koi 1894.01 and koi 2133.01 .
one normally expects a transit duration of several hours for period such as these .
however , the values of @xmath15 and @xmath16 shown in this table imply a @xmath17 larger than 20 hours ! this does not appear to make sense until one considers the stellar radius .
note from equation [ duration ] that , for an edge - on orbit and small @xmath6 , the transit duration scales linearly with the size of the star .
for these two candidates , the stellar radii are 8.6 and 9.3 solar radii respectively thus resulting in a large @xmath16 and a significant eccentricity required to be consistent with observations .
note , however , that we have assumed @xmath18 .
as one increases the impact parameter , the predicted transit duration will decrease and thus become closer to its measured value .
results for individual cases extracted from the global distribution , such as those in table [ ecctab ] , must therefore be treated with caution .
we mentioned in section [ kepler ] that the analysis of the kepler objects included only candidates for which @xmath19 .
here we perform a separate study by repeating the calculations of @xmath16 for the kepler candidates for a range of planetary radii .
we allow the candidate sample to include all radii larger than @xmath20 to @xmath21 and calculate the mean of the @xmath16 distribution in each case .
we show our results in figure [ mean ] . for the kepler candidates as a function of minimum planetary radius included in the sample ( [ correlation]).,width=309 ]
an interpretation of this figure is that the eccentricity distribution of exoplanets remains relatively flat until we probe below planets the size of neptune . at that point
the eccentricity distribution of the orbits becomes rapidly and significantly more circular .
this is not unexpected since we understand from the solar system that the inefficiency of tidal dissipation ( the quality factor @xmath22 ) is much larger for high - mass than for low - mass planets @xcite , resulting in shorter tidal circularization time scales for smaller planets .
one aspect of the kepler candidate sample that may influence this result is that they are dominated by planets at smaller semi - major axes since these have much larger transit probabilities . as indicated by @xcite and @xcite ,
multi - planet systems comprise a large proportion of the total kepler candidate sample and these systems in particular are less prone to be false - positives .
the findings that planet occurrence increases with decreasing planet mass ( see for example * ? ? ?
* ) then suggests that smaller planets find stable architectures in systems with circular orbits and without large planets in eccentric orbits .
this lends credence to two scenarios : ( 1 ) core accretion forming terrestrial planets in circular orbits , and ( 2 ) disk instability and capture scenario explaining the existence of giant planets in eccentric orbits .
a potential alternative explanation for the dependence of @xmath16 upon @xmath23 is a correlation between planetary radii and semi - major axes in the kepler sample due to completeness .
smaller planets in closer orbits would be more likely to have had their orbits circularized leading to their domination of the sample .
we do not , however , see such an effect but instead find no correlation between planet size and period in the kepler catalog for the selected range of orbital periods .
thus , we conclude that this effect is not the cause of the observed radius / eccentricity correlation .
there are various sources of potential systematic noise inherent in the data used to perform this analysis .
for example , we have not taken the stellar limb - darkening into account when considering transit durations . however , since we only consider the total transit duration ( first contact to last contact ) , this will be a negligible effect . for the kepler candidates , the primary source of uncertainty arises from the stellar parameters that are used to derive many of the planetary candidate parameters .
the primary difficulties arise from the stellar radii whose precision is usually worse than @xmath2410% , even when spectra are available .
our assumption is that these uncertainties are not significantly biased in one direction of the other and thus only add white noise to the overall statistical properties .
one method to test this assumption is to consider multi - candidate systems for which a change in the stellar radius will effect all calculated transit durations in a similar way .
for example , koi 157 ( kepler-11 ) has 6 detected planet candidates with measured transit durations that are shorter than predicted , and with an estimated host star radius of 1.06 @xmath25 .
three of those discrepancies are quite small making them almost consistent with a circular orbit , as noted by @xcite . attempting to force circular orbits by reducing the stellar radius
slightly makes 2 of the planets consistent with a circular orbit but leaves the other durations highly discrepant .
a more global test of the assumption is the application of a range of uniform scaling factors to the stellar radii to try to produce transit durations consistent with circular orbits for the bulk of the distribution .
these corrections did not change the distribution shown in figure [ keplercand ] , leading to the conclusion that our calculated @xmath16 values are not affected by systematically incorrect stellar radii . considering the uncertainties in the radii of the kepler host stars , our limits on the eccentricities of the specific kepler candidates discussed in section [ kepler ]
should be treated in that context . indeed ,
this is why we concentrate our comments on the global distribution of all the kepler objects and their parameters , rather than on individual cases .
furthermore , the kepler sample is only complete to @xmath26 au with declining completeness beyond this to @xmath27 au .
thus , the number of candidates beyond 0.5 au is smaller but still sufficient for a valid comparison to be made .
one aspect of the kepler candidates that was not taken into account was the multiplicity of the systems .
as suggested at the conclusion of the previous section , the multiplicity may indeed play a significant role in stabilizing planets in approximately circular orbits , particularly for those in the low mass / size regime .
this is true of the radial velocity planets also , some of which are known to lie in multiple systems of super - earth mass planets and with relatively circular orbits such as the system orbiting hd 10180 @xcite . in section [ intro ]
we mentioned the eccentricity bias found by @xcite due to low - amplitude signals in rv samples .
this is a small but real effect which depends upon the sampling rate and has the consequence of under - estimating the number of near - circular orbits .
@xcite develop a figure - of - merit and find that only @xmath28% of the planets considered in their sample are affected by this bias .
the samples studied here are two small to detect such an effect , but we mention it here as a consideration for future similar work for which the sample sizes and , more particularly , the period range explored have grown substantially . finally , a minor impact on the stellar radii that should be noted is the one due to the relation between planet frequency and stellar metallicity .
@xcite explored the mass - metallicity relationship for stars that harbor planets and found a positive correlation of planet frequency with both stellar mass and metallicity , in accordance with the findings of @xcite .
this positive correlation was also found empirically for m dwarfs by @xcite . for a given stellar mass ,
a larger metallicity leads to a smaller radius in order to reach hydrostatic equilibrium .
the implication for this study is that many of the kepler host stars will have relatively high metallicity leading to an over - estimated radius .
however , this effect is at the level of a few percent and not expected to interfere with the results of this study .
by conducting a transit survey that is sensitive to long enough periods , it is expected that one will eventually reproduce the eccentricity distribution found amongst radial velocity planets .
this has not been possible until very recently , when the kepler presented a large sample of long - period planets candidates , providing the incentive for this study and a similar one by @xcite . for individual planets ,
the eccentricity may be discerned via asymmetry in the shape of ingress and egress @xcite .
this requires exquisite photometry and is highly sensitive to @xmath10 .
we have shown here the consistency of the kepler candidates eccentricity distribution with their rv planets counterparts .
the correlation of eccentricity with planet size is also an expected result based upon the discoveries of small planets in multiple systems and indicates that there is an empirical approach from which to both reverse - engineer formation scenarios and predict future stability patterns .
the authors would like to thank rory barnes , eric ford , andrew howard , and suvrath mahadevan for several useful discussions .
we would also like to thank the anonymous referee , whose comments greatly improved the quality of the paper .
this research has made use of the exoplanet orbit database and the exoplanet data explorer at exoplanets.org .
this research has also made use of the nasa exoplanet archive , which is operated by the california institute of technology , under contract with the national aeronautics and space administration under the exoplanet exploration program .
barnes , j.w . , 2007 , pasp , 119 , 986 batalha , n.m . , et al . , 2012 , apjs , submitted ( arxiv:1202.5852 ) borucki , w.j . , et al . , 2011 , apj , 728 , 117 borucki , w.j . , et al . , 2011 , apj , 736 , 19 butler , r.p . , et al . , 2006 , apj , 646 , 505 burke , c.j . , 2008 , apj , 679 , 1566 fischer , d.a . , valenti , j. , 2005 , apj , 622 , 1102 ford , e.b . , quinn , s.n . , veras , d. , 2008 , apj , 678 , 1407 ford , e.b . ,
rasio , f.a . , 2008 ,
apj , 686 , 621 goldreich , p. , soter , s. , 1966 , icarus , 5 , 375 howard , a.w . , et al . , 2010 , science , 330 , 653 johnson , j.a . ,
aller , k.m . ,
howard , a.w . , crepp , j.r . , 2010 , pasp , 122 , 905 juri , m. , tremaine , s. , 2008 , apj , 686 , 603 kane , s.r .
, gelino , d.m . , 2012 ,
pasp , 124 , 323 kane , s.r . , von braun , k. , 2008 , apj , 689 , 492 kane , s.r . , von braun , k. , 2009 , pasp , 121 , 1096 kipping , d.m . , 2008 , mnras , 389 , 1383 lissauer , j.j . ,
et al . , 2011 , nature , 470 , 53 lissauer , j.j .
, et al . , 2011 , apjs , 197 , 8 lissauer , j.j . ,
et al . , 2012 , apj , 750 , 112 lovis , c. , et al . , 2011 ,
a&a , 528 , 112 malmberg , d. , davies , m.b . , 2009 , mnras , 394 , l26 mandel , k. , agol , e. , 2002 , apj , 580 , l171 matsumura , s. , takeda , g. , rasio , f.a . , 2008 , apj , 686 , l29 moorhead , a.v . , et al . , 2011 , apjs , 197 , 1 plavchan , p. , bilinski , c. , currie , t. , 2012 , apjl , submitted ( arxiv:1203.1887 ) pont , f. , husnoo , n. , mazeh , t. , fabrycky , d. , 2011 , mnras , 414 , 1278 seager , s. , malln - ornelas , g. , 2003 , apj , 585 , 1038 shen , y. , turner , e.l .
, 2008 , apj , 685 , 553 smalley , b. , 2005 , mem .
, 8 , 130 terrien , r.c . , mahadevan , s , bender , c.f . ,
deshpande , r. , ramsey , l.w . ,
bochanski , j.j . , 2012 ,
apj , 747 , l38 von braun , k. , kane , s.r . ,
ciardi , d.r .
, 2009 , apj , 702 , 779 wang , j. , ford , e.b . , 2011 , mnras , 418 , 1822 wright , j.t . ,
et al . , 2011 , pasp , 123 , 412 zakamska , n.l . , pan , m. , ford , e.b .
, 2011 , mnras , 410 , 1895 | the eccentricity distribution of exoplanets is known from radial velocity surveys to be divergent from circular orbits beyond 0.1 au .
this is particularly the case for large planets where the radial velocity technique is most sensitive .
the eccentricity of planetary orbits can have a large effect on the transit probability and subsequently the planet yield of transit surveys .
the kepler mission is the first transit survey that probes deep enough into period - space to allow this effect to be seen via the variation in transit durations .
we use the kepler planet candidates to show that the eccentricity distribution is consistent with that found from radial velocity surveys to a high degree of confidence .
we further show that the mean eccentricity of the kepler candidates decreases with decreasing planet size indicating that smaller planets are preferentially found in low - eccentricity orbits .
planetary systems techniques : photometric radial velocities |
the main prediction of the quantum state of the universe is the emergence of classical universe that is a manifest fact of the observable universe .
therefore , predicting classical cosmology is a constraint on the theory of the state @xcite . in minisuperspace models of cosmology , the classical gravitational field equations are
often non - linear and the quantization procedure is not unique .
there are infinite number of transformations that recast such equations into forms with different finite number of degree of freedom interpretations .
distinctly an appropriate transformation , state and quantization should be those that have some chance of yielding a classical limit not too far removed the original classical predictions @xcite . by canonical quantization of gravity ,
one finds various defects when one inspects its content in detail .
first , since the canonical hamiltonian of gravity is written as a linear combination of the constraints , it annihilates the physical quantum state . hence , the time evolution is lost from theory . besides a new problems
aries when one tries to apply the quantum theory of gravity to cosmology : when our working tool is the wave function , which should be obtained by wheeler - dewitt ( wdw ) equation or path integral , we need to know how it is possible to construct an adequate wave packet that would peaked around the original classical cosmological model @xcite . in ordinary circumstances ,
the wave packet reduction in the copenhagen interpretation gives rise to no practical problem as far as we regard the quantum mechanics as describing the dynamics of an ensemble of identical systems .
however , in cosmology , there is only one universe as a system .
therefore , it is not clear how the state of the universe has any well - defined wave function .
also , in quantum cosmology the observer itself is an element of universe . in the standard interpretation of quantum mechanics , during an observation , the quantum system must interact with a classical domain ( measuring instrument , observer body , ... ) . in von neumann s view , the necessity of a classical domain comes from the way it solves the measurement problem @xcite . in a conventional impulsive measurement , where the coupling interaction between the measured system and the classical measuring devise is of short duration and strong
, the wave function plus measuring devise splits into many branches which do not overlap , each one containing the measured system in an eigenstate of the measured observable , and the `` pointer '' of the measuring devise pointing to the corresponding eigenvalue .
however , in the end of the measurement processes , the observer measures only one eigenvalue and the immediate repeating of measurement gives the same result .
therefore , the wave function collapses into an eigenstate of the observable that is registered and the other branches disappear .
but , a real collapse of wave function can not be described by the unitary quantum evolution .
therefore , the copenhagen interpretation needs to assume the existence of fundamental process in a measurement which occur outside the quantum system , in a classical domain .
it is obvious that in quantum cosmology , as a quantum theory of whole universe , there is no place for classical domain outside of it .
hence , it seems that the copenhagen interpretation can not be applied to quantum cosmology or it should be improved by means of further concepts .
one possibility is invoking environmental decoherence @xcite .
however , quantum decoherence is not yet a complete answer to the measurement problem @xcite .
it dose not explain the apparent collapse after the measurement is completed , or why all but one of the diagonal elements of the density matrix become null when the measurement is finished @xcite . also , in its developments like the consistence histories approach @xcite the important role played by observers is not yet explained @xcite and it is not clear how to describe a quantum universe when the background geometry is not classical @xcite . nevertheless , there are some alternative solutions to the above quantum cosmological difficulties , which together with decoherence may solve the measurement problem maintaining the universality of quantum theory and emergence of classical universe . in this line
we should cite the de broglie - bohm interpretation of quantum cosmology @xcite , quantum hamilton - jacobi cosmology @xcite and deformation quantization of cosmology @xcite .
the advantage of deformation quantization is that it makes quantum cosmology look like the hamiltonian formalism of cosmology .
in other words , deformation quantization can be viewed as a deformation of the structure of the algebra of classical observables , rather than a radical change in the nature of the observables .
this is done by avoiding the operator formalism @xcite .
deformation quantization , which is presented as weyl - wigner - groenewold - moyal phase space quantization , describes a quantum system in terms of the classical @xmath0-number variables @xcite .
this means that operators are mapped into the @xmath0-number functions so that their compositions could be obtained by the star product that is noncommutative but associative .
therefore , the observables would be classical functions of the phase space .
quantum structure is constructed by replacing pointwise products of classical observables of the phase space @xmath1 , by star product @xcite .
the product of two smooth functions , say @xmath2 and @xmath3 , on a poisson - moyal manifold is given by @xmath4 where @xmath5 plays the role of the deformation parameter . the first term @xmath6 denotes the common pointwise product of @xmath7 and @xmath8 .
also , the coefficients @xmath9 are bidifferential operators , where their product is noncommutative @xcite .
these coefficients satisfy the following properties @xmath10 where @xmath11 denotes the ordinary poisson bracket . in equations ( [ 0 - 2 ] )
, the first expression means that in the limit , @xmath12 , the star product of @xmath7 and @xmath8 agrees with the pointwise products of these two functions in classical phase space .
the second expression shows that at the lowest order of the deformation parameter , the @xmath13-commutator @xmath14_*=f*g - g*f$ ] tends to the poisson bracket . in flat spaces
, there is a special star product which has long been known . in this case , the components of the poisson tensor can be considered constant and consequently it is possible to define the following moyal star product @xcite @xmath15 the last equality in ( [ moyal product ] ) suggests that for a moyal star product @xmath16 of two functions @xmath17 and @xmath18 the weyl - wigner correspondence reads @xmath19 where now @xmath20 should be understood as an operator acting on @xmath21 .
according to this prescription we have to replace the position and momentum variables in @xmath17 by pseudo - operators containing position and momentum and their derivatives , that is @xmath22 instead of the usual correspondence @xmath23
. these pseudo - operators sometimes carry the name bopp pseudo - operators
. these pseudo - operators act , not on functions defined on @xmath24 as ordinary weyl operators do , but on functions ( or distributions ) defined on the noncommutative phase space @xmath25 .
in fact , bopp pseudo - differential operators is a tool of choice for the study of deformation quantization which it reduces to a weyl calculus of a particular type .
also , equation ( [ new1 ] ) shows that noncommutative quantum mechanics can also be reduced to bopp calculus from an operator point of view .
one of the most important components of deformation quantization is the wigner quasiprobability distribution function @xcite .
in fact , it is a generating function for all spatial autocorrelation functions of a given quantum mechanical wave function @xcite .
the relation of wigner function with wave function of system @xmath26 in a @xmath27-dimensional phase space is @xmath28 in this formalism of quantum mechanics , expectations of observables and transition amplitudes are phase space integrals of @xmath0-number functions , weighted by the wigner function , as in statistical mechanics . in the next section
we will investigate the quantum cosmology of a flat friedmann - lematre - robertson - walker ( flrw ) universe , filled with radiation plus dust or cosmic string .
using wigner function we will show that the deformed cosmology predicts a good correlation with the corresponding classical cosmology .
natural units @xmath29 are used throughout the paper .
there exist two fundamentally different approaches to quantization of general relativity : the particle physics programme and the canonical quantum gravity programme . in the first programme ,
the basic entity is the graviton , the quantum of the gravitational field . such a particle is deemed to propagate in a background minkowski spacetime , @xmath30 , and like all elementary particles , is associated with a specific representation of the poincar group . by fixing the background topology and differential structure of spacetime manifold to be that of minkowski spacetime , and then splinting spacetime into background , @xmath31 , and dynamical , @xmath32 , parts , the field @xmath32 is quantized by the standard field methods under the assumption that the gravitational interaction , like other standard matter interactions , involves the exchange of gravitons . on the other hand ,
the canonical approach to quantum gravity starts with a reference foliation of spacetime with respect to which the appropriate canonical variables are defined . in this sense ,
quantization of gravitation is quantization of the metrical structure of spacetime , which satisfies a dynamical principle and dynamical equations .
the essence of this approach is expressed by the name quantum geometrodynamics . to study the nature of the difficulties of quantum geometrodynamics
, it is convenient to use a particular minisuperspace approximation .
the model nature of these studies is a consequence of various factors , of which the main are : 1 ) a minisuperspace of definite symmetry is selected and it is assumed that the symmetry is preserved in the quantization process .
2 ) to maintain a definite symmetry of minisuperspace , one requires a material source ( perfect fluid , yang - mills fields , scalar fields or spinor fields ) , which is also quantized when gravitation is quantized .
it is clear that in the domain of quantum geometrodynamics ( planck scale ) the description of the material source will differ from the one adopted in quantum field theory , and therefore in quantum geometrodynamics the source is taken into account only formally by the addition of new degrees of freedom to the equations that describe the dynamical geometry @xcite . due to the quantum nature of the model , as a first approximation , the matter content should be described by some sorts of fields , as done in @xcite .
however , general exact solutions are hard to find at the presence of yang - mills , scalar or spinor fields and the hilbert space structure is obscure and it is a subtle matter to recover the notion of a semiclassical time @xcite . in addition , since the bekenstein bound tells us that the information content of the very early universe is zero , the only physical variable we have to take into account is the scale factor of the universe , and the density and pressure of the matter field .
hence , we only have to quantize the flrw universe for a perfect fluid @xcite .
also , perfect fluid has the advantage of introducing a variable which can naturally be identified with time , leading to a well - defined hilbert space structure @xcite .
another attractive feature of the phenomenological perfect fluid description of matter degree of freedom is that it allows us to treat the barotropic equation of state which allows us to obtain general exact solutions .
the line element of spatially flat flrw universe is @xmath33 where @xmath34 is the lapse function and @xmath35 is the scale factor .
the action functional that consists of a gravitational part and a matter part when the matter field is considered as a perfect fluid is given by @xcite @xmath36 where @xmath8 is the determinant of spacetime metric and @xmath37 is the ricci scalar and @xmath38 is the total energy density .
we assume a universe filled with noninteracting perfect fluids with energy densities of @xmath39 where @xmath40 denotes the equation of state parameter of @xmath41-th component of fluid and @xmath42 is the energy density at the measuring epoch .
the action ( [ action1 ] ) reduce to @xmath43 where @xmath44 is the spacial volume of 3-metric .
let us rewrite the action ( [ action2 ] ) in terms of measurable quantities in cosmology .
this will help us to compare the quantum cosmological model with the corresponding classical model .
first , we define a new lapse function by @xmath45 . by writing the energy density of various components of fluid in terms of corresponding density parameters , @xmath46 ( @xmath47 is the hubble parameter at the measuring epoch )
the energy densities will be @xmath48 .
also if we use a new dimensionless scale factor defined by @xmath49 and a new dimensionless time coordinate by @xmath50 , the lagrangian of model in conformal frame up to a multiplicative constant @xmath51 , will be @xmath52 where over dot denotes differentiation respect to @xmath30 .
the conjugate momentum to the scale factor , @xmath53 , and the primary constraint are given by @xmath54 consequently , the hamiltonian corresponding to lagrangian ( [ lagrangian ] ) will be @xmath55 in hamiltonian ( [ canonical hamilton ] ) , @xmath56 is a lagrange multiplier ; therefore , it enforces the hamiltonian constraint @xmath57 where @xmath58 denotes the super - hamiltonian .
the super - hamiltonian at the initial time @xmath59 reduces to well - known relation between density parameters @xmath60 .
the deformation quantization of this simple model is accomplished straightforwardly by replacing the ordinary products of the observables in phase space by the moyal product .
therefore , hamiltonian constraint ( [ constraint ] ) becomes the moyal - wheeler - dewitt ( mwdw ) equation by replacing the classical hamiltonian ( [ constraint ] ) with its deformed counterpart @xcite @xmath61 since the @xmath62-product involves exponentials of derivative operators , it may be evaluated in practice through translation of function arguments ( see , eq.([new1 ] ) ) .
therefore , the mwdw equation ( [ sh2 ] ) is equivalent to @xmath63 let us now first investigate the classical - quantum correlation in a radiation - dust filled universe . in this case
, the classical super - hamiltonian ( [ constraint ] ) will be @xmath64 where @xmath65 and @xmath66 denote the density parameters of dust and radiation respectively , obeying relation @xmath67 at the measuring epoch , @xmath68 .
also , the mwdw equation ( [ b1 ] ) will be @xmath69 note that we should order the kinetic term as @xmath70 , where @xmath71 takes into account the factor ordering ambiguity .
this is equivalent to the factor ordering in corresponding wdw equation given by @xmath72 @xcite . in this paper
we will consider the choice @xmath73 ordering which is equivalent to the laplace - beltrami operator of conformal frame in the corresponding wdw equation .
the wigner function is real , hence by separation the real and imaginary parts of mwdw equation ( [ sh4 ] ) we obtain two coupled partial differential equations @xmath74 the first equation does not involves the partial derivatives of @xmath75 .
however , the second phase space equation enforces a special symmetry on the solutions .
the solution of second partial differential equation of ( [ sh5 ] ) is , @xmath76 , where @xmath7 denotes a general real function . with the help of definition of new variable @xmath77 and the relation @xmath78
following from the chain rule , the first partial differential equation of ( [ sh5 ] ) reduces to following second order ordinary differential equation @xmath79 the finite value solution of this equation is @xmath80 where @xmath81 and @xmath82 denotes the airy function of first kind .
the locus of extremums of the above wigner function is the following deformed super - hamiltonian @xmath83 where @xmath84 are the zeroes of derivative airy functions @xmath85 .
see table ( 1 ) for the first several terms of @xmath84 sequences .
.negatives of zeroes of @xmath86 for @xmath87 . [ cols="<,^,^,>,>,>,>,>,>,>",options="header " , ] [ t2 ] the modified super - hamiltonian ( [ re1 ] ) represents a universe filled with cosmic string and radiation fluids , where the density parameter of radiation is given by @xmath88 note that the radiation part of ( [ re1 ] ) has totally quantum origin and the corresponding classical universe is filled only with the cosmic string fluid . in the late times , where the universe is cosmic string dominated , the predictions of both models are the same . but for very small values of scale factor , the emerged universes for various values of @xmath89 are non - singular and radiation dominated .
the first zero of derivative bessel function is zero , consequently for @xmath90 the prediction of quantum cosmology is a cosmic string filled singular universe , same as the corresponding classical universe ( [ sh7 ] ) .
fig.(2 ) shows the corresponding wigner function with corresponding classical trajectory in phase space . for large values of scale factor ,
@xmath53 , there is a good correlation between the quantum quasiprobability distribution and the classical trajectory in phase space .
in this paper we studied the deformation or phase space quantization of a flat quantum flrw model , whose matter is either a fluids of radiation plus dust , or cosmic string .
we show that the peaks of wigner quasiprobability distribution function of quantum universe filled with cosmic string are coincide with the emerged universes whose are filled with cosmic string and radiation fluids . also , for a universe filled by the radiation and dust fluids , the peaks of wigner quasiprobability distribution function are coincide with the same classical universes but with modified density parameter of radiation .
consequently the behaviour of emerged quantum universes are different with the corresponding classical models for the very small values of scale factors , where the quantum universes are non - singular . on the other hand , for large values of scale factor , the emerged quantum universes are coincide with corresponding classical universes .
this behavior can be interpreted that the classical approximation of the universe becomes better and better as the universe expands .
we believe the above results offer an insight into the relation between classical and quantum cosmologies and that the particular simple models that we have studied may serve as useful starting point for more ambitious investigation .
we are aware that our results are obtained within a very simple as well as restricted setting .
nevertheless , we think they are intriguing and provide motivation for subsequent research works .
a wider analysis , with less restrictive cosmologies and/or other matter fields , should follow .
the presence of phenomenological fluid matter was broadly used in , e.g. , @xcite so that exact solutions of the ( simplified ) wdw equation could be obtained . using instead , e.g.
, a scalar , spinor or yang - mills fields would be more generic and more realistic from the point of view of matter interaction with the gravitational field in a high energy regime , where quantum effects can be expected .
we are leaving the above enticing research lines for future works .
hartle , s.w . hawking and t. hertog , phys .
d * 77 * , 123537 ( 2008 ) , [ arxiv:0803.1663 ] . t. dereli , m. nder and robin w. tucker , class . quantum grav . * 10 * , 1425 ( 1993 ) .
alvarenga , j.c .
fabris , n.a .
lemos and g.a .
monerat , gen .
. grav . * 34 * , 651 ( 2002 ) , [ arxiv : gr - qc/0106051 ] ; n.a .
lemos , j. math . phys . * 37 * , 1449 ( 1996 ) ; g.a .
monerat , g. oliveira - neto , e.v .
corra silva , l.g .
ferreira filho , p. romildo jr .
fabris , r. fracalossi , f.g .
alvarenga and s.v.b .
gonalves , phys .
d * 76 * , 024017 ( 2007 ) ; f.g .
alvarenga , a.b .
batista and j.c .
fabris , int . j. mod
d * 14 * , 291 ( 2005 ) , [ arxiv : gr - qc/0404034 ] ; p. pedram and s. jalalzadeh , phys .
b * 659 * , 6 ( 2008 ) , [ arxiv:0711.1996 ] ; p. pedram , s. jalalzadeh and s.s .
gousheh , phys . lett .
b * 655 * , 91 ( 2007 ) , [ arxiv:0708.4143 ] ; p. pedram and s. jalalzadeh , phys .
d * 77 * , 123529 ( 2008 ) , [ arxiv:0805.4099 ] .
r. omns , _ the interpretation of quantum mechanics _ , princeton university press , princeton , ( 1994 ) .
zeh , found .
* 1 * , 69 ( 1970 ) ; e. joos and h.d .
zeh , z. phys .
b * 59 * , 223 ( 1985 ) ; w.h .
zurek , phys . today * 44 * , 36 ( 1991 ) .
mukhanov , in _ physical origins of time asymmetry _ , ed . by j.j .
halliwell , j. prez - mercader and w.h .
zurek , cambride university press ( 1994 ) ; h.d .
zurek , in _ decoherence and the appearance of classical world in quantum theory _
springer - verlag , berlin , ( 1996 ) .
n. pinto - neto , brazilian j. phys .
* 30 * , 330 ( 2000 ) . m. gell - mann and j.b .
hartle , in _ complexity , entropy and the physics of information _ , ed . by w.h .
zurek , addison - wesley , ( 1990 ) .
paz and w.h .
zurek , phys .
d * 48 * , 2728 ( 1993 ) .
n. pinto - neto and j.c .
fabris , class .
quantum grav . * 30 * , 143001 ( 2013 ) , [ arxiv:1306.0820 ] .
f. shojai and a. shirinifard , int . j. mod
d * 14 * , 1333 ( 2005 ) , [ arxiv : gr - qc/0504138 ] ; p. pedram and s. jalalzadeh , phys . lett .
b * 660 * , 1 ( 2008 ) , [ arxiv:0712.2593 ] ; p. peter and n. pinto - neto , phys .
d * 78 * , 063506 ( 2008 ) , [ arxiv:0809.2022 ] . m. fathi , s. jalalzadeh and p.v .
moniz , eur .
j. c * 76 * , 527 ( 2016 ) , [ arxiv:1609.04488 ] .
f. antonsen , [ arxiv : gr - qc/9712012 ] ; phys .
d * 56 * , 920 ( 1997 ) ; h. quevedo and j.g .
tafoya , gen .
37 * , 2083 ( 2005 ) [ arxiv : gr - qc/0401088 ] ; h. garca - compen , f.j .
turrubiates , int . j. mod .
a * 26 * , 5241 ( 2011 ) ; r. cordero , h. garca - compen and f.j .
turrubiates phys .
d * 83 * , 125030 ( 2011 ) , [ arxiv:1102.4379 ] ; m. rashki and s. jalalzadeh , phys .
d * 91 * , 023501 ( 2015 ) , [ arxiv:1412.3950 ] .
j. moyal , proc .
cambridge philos .
* 45 * , 99 ( 1949 ) .
h. weyl , z. phys .
* 46 * , 1 ( 1927 ) ; e. wigner , phys .
rev . * 40 * , 749 ( 1932 ) ; h. j. groenewold , physica ( utrecht ) * 12 * , 405 ( 1946 ) ; f. bayen , m. flato , c. fronsdal , a. lichnerowicz , and d. sternheimer , ann . phys .
( n.y . ) * 111 * , 111 ( 1978 ) .
m. kontsevich , lett .
* 48 * , 35 ( 1999 )
, [ arxiv : math/9904055 ] . a.c .
hirschfeld and p. henselder , am .
* 70 * , 537 ( 2002 ) , [ arxiv : quant - ph/0208163 ] .
m hillery , r. f. oconnell , m. c. scully , and e. p. wigner , phys . rep . * 106 * , 121 ( 1984 ) ; t. curtright , d. fairlie , and c. zachos , phys .
d * 58 * , 025002 ( 1998 ) ; m. levanda and v. fleurov , ann . phys .
* 292 * , 199 ( 2001 ) ; c. zachos , int
a * 17 * , 297 ( 2002 ) .
b. v. fedosov , j. differential geom .
* 40 * , 213 ( 1994 ) ; m. gadella , fortschr . phys . * 43 * , 229 ( 1995 ) .
v.g . lapchinskii and v.a .
rubakov , theor .
* 33 * , 1076 ( 1977 ) . c. kiefer , phys .
d * 38 * , 1761 ( 1988 ) .
isham , [ arxiv : gr - qc/9210011 ] .
tipler , rept .
. phys . * 68 * , 897 ( 2005 ) , [ arxiv:0704.3276 ] .
gotay and j. demaret , phys .
d * 28 * , 2402 ( 1983 ) ; f.g .
alvarenga , j.c .
fabris , n.a .
lemos and g.a .
monerat , gen .
. grav . * 34 * , 651 ( 2002 ) , [ arxiv : gr - qc/0106051 ] ; p. pedram , m. mirzaei , s. jalalzadeh and s. s. gousheh , gen .
. grav . * 40 * , 1663 ( 2008 ) , [ arxiv:0711.3833 ] ; b. vakili , class .
* 27 * , 025008 ( 2010 ) , [ arxiv:0908.0998 ] ; p. pedram , s. jalalzadeh and s.s .
gousheh , phys .
b. * 655 * , 91 ( 2007 ) , [ arxiv:0708.4143 ] ; p. pedram , s. jalalzadeh and s.s .
gousheh , class .
* 24 * , 5515 ( 2007 ) , [ arxiv:0709.1620 ] .
s.w . hawking and g.f.r .
ellis , _ the large scale structure of space - time _
( cambridge university press , cambridge , united kingdom , 1973 ) .
r. steigl and f. hinterleitner , class .
quantum grav .
* 23 * , 3879 ( 2006 ) , [ arxiv : gr - qc/0511149 ] .
m. hillery et al . , phys . rep .
* 106 * , 121 ( 1984 ) .
kung , gen .
* 27 * , 35 ( 1995 ) , [ arxiv : hep - th/9302016 ] ; j. demaret , gen .
. gravit . * 11 * , 453 ( 1979 ) ; t. banks , w. fischler and l. mannelli , phys .
d * 71 * , 123514 ( 2005 ) [ arxiv : hep - th/0408076 ] ; b. majumder , int .
d * 22 * , 1350079 ( 2013 ) , [ arxiv:1307.5273 ] ; d. atkatz and h. pagels , phys . rev .
d * 25 * , 2065 ( 1982 ) ; p.s .
letelier and j.p.m .
pitelli , phys .
d * 82 * , 104046 ( 2010 ) , [ arxiv:1010.3054 ] ; b. majumder and n. banerjee , gen .
. gravit . * 45 * , 1 ( 2013 ) , [ arxiv:1208.4481 ] ; s. nojiri and s.d .
odintsov , phys .
b * 562 * , 147 ( 2003 ) , [ arxiv : hep - th/030311 ] . | in this paper we study the quantum cosmology of homogeneous and isotropic cosmology , via the weyl - wigner - groenewold - moyal formalism of phase space quantization , with perfect fluid as a matter source .
the corresponding quantum cosmology is described by the moyal - wheeler - dewitt equation which has exact solutions in moyal phase space , resulting in wigner quasiprobability distribution functions peaking around the classical paths for large values of scale factor .
we show that the wigner functions of these models are peaked around the non - singular universes with quantum modified density parameter of radiation .
+ + keywords : quantum cosmology , moyal - wheeler - dewitt equation , deformation quantization |
rotating magnetospheres are widely believed to be responsible for the relativistic jet phenomenon in active galactic nuclei ( agn ) @xcite .
here we adress the question whether centrifugal acceleration of charged test particles at the base of such a jet magnetosphere may possibly produce a seed population of relativistic electrons which is required for efficient particle acceleration . for , in order to explain the origin of the nonthermal emission extending up to tev energies in some blazars ,
several acceleration processes have been proposed among which fermi - type particle acceleration mechanisms ( i.e. diffusive shock acceleration @xcite ) are quite promising .
however such kind of mechanisms require a pre - accelerated seed population of electrons with lorentz factors of the order of @xmath5 @xcite .
it seems therefore quite interesting whether in the case of agn centrifugal acceleration by rotating jet magnetosphere may potentially fill this gap by providing pre - accelerated seed particles . for an analytical treatment ,
we consider the following simplified model : motivated by mhd - scenarios for the origin of jets via rotating jet magnetospheres @xcite ( see fig . [ jet ] ) a projected two - dimensional model topology is applied where the magnetic field is supposed to rotate rigidly with a fraction of the rotational velocity of the black hole @xcite .
test particles with rest mass @xmath6 and charge @xmath7 are assumed to be injected at time @xmath8 and position @xmath9 with velocity @xmath10 parallel to the rotating field line .
consider the forces acting on a particle in a rotating frame of reference @xcite : particles , which are injected at ( @xmath8,@xmath9 ) with velocity @xmath10 along the magnetic field line @xmath11 experience a centrifugal force in the radial direction given by @xmath12 where @xmath13 denotes the lorentz factor and @xmath14 the angular velocity of the field .
additionally , there is also a relativistic coriolis term in the noninertial frame governed by the equation @xmath15 which acts as a deviation - force in the azimuthal direction . in the inertial rest frame
the particle sees the field line bending off from its initial injection position , therefore it experiences a lorentz force ( @xmath16 ) @xmath17 where @xmath18 is the relative velocity between the particle and the magnetic field line . due to the lorentz force
a particle tries to gyrate around the field line .
initially , the direction of the lorentz force is perpendicular to the direction of the coriolis force , but as a particle gyrates , it changes the direction and eventually becomes antiparallel to the coriolis force .
hence , the bead - on - the - wire approximation is valid if the lorentz force is not balanced by the coriolis force @xcite . in this case , the accelerated motion of the particle s guiding center due to the centrifugal force may be written as @xmath19 where @xmath20 .
the constrained motion is then given by the azimuthal components of forces @xmath21 generally , the bead - on - the - wire approximation is supposed to break down if @xmath22 exceeds @xmath23
( i.e. when @xmath24 in eq .
[ constraint ] becomes @xmath25 ) .
using the argument that the hamiltonian for a bead on a relativistically moving wire @xmath26 is a constant of motion , the equation for the radial accelerated motion could be reduced to a simple form which has been solved analytically yielding @xcite @xmath27 where @xmath28 , ( @xmath29 ) is the jacobian elliptic cosine ( sine , respectively ) , and @xmath30 is an elliptic integral of the first kind , i.e. @xmath31 with @xmath32 .
the lorentz factor may then be written as @xmath33^{2}}\,,\ ] ] or , if expressed as a function of the radial co - ordinate , as @xmath34 apart from radiation losses ( e.g. inverse - compton losses in the radiation field of the accretion disk , see @xcite ) , the maximum attainable lorentz factor @xmath1 is in particular limited by the breakdown of the bead - on - the - wire approximation ( i.e. when the particle leaves the field line and thus , acceleration becomes ineffective ) in the vicinity of the light cylinder @xmath0 . using the definition of the hamiltonian @xmath35 and eq .
[ gamma_r ] and setting @xmath36 , one may derive an upper limit for the maximum lorentz factor @xmath1 from eq .
[ constraint ] @xmath37 where @xmath38 denotes the magnetic field strength at the light cylinder and where for clarification @xmath39 has now been inserted . for typical bl lac conditions , i.e. a light cylinder radius @xmath40 m , and a field strength @xmath41 t , eq .
[ gmax ] results in an upper limit on the maximum lorentz factor @xmath42 .
the results derived in the simple toy - model presented here support flares on accretion disks as providing a seed population of relativistic electrons with lorentz factors up to @xmath43 in bl lac type objects .
such pre - accelerated particles are required for models involving diffusive shock acceleration of @xmath44 in relativistic jets , cf .
@xcite , @xcite .
particle acceleration by rotating jet magnetospheres may thus possibly represent an interesting explanation for the required pre - acceleration .
begelman , m.c .
, `` magnetic propulsion of jets in agn , '' in _ the nature of compact objects in active galactic nuclei _ , edited by a. robinson , and r. terlevich , univ .
press , cambridge , 1994 , pp .
361 - 367 .
blandford , r.d . , and payne , d.g . , _ mnras _ * 199 * , 883 ( 1982 ) .
camenzind , m. , `` stationary relativistic mhd flows , '' in _ solar and astrophysical magnetohydrodynamic flows _ , edited by k.c .
tsinganos , kluwer academic publ . ,
dordrecht , 1996 , pp .
699 - 725 .
drury , l.oc .
phys . _ * 46 * , 973 ( 1983 ) .
fendt , c. , _ a&a _ * 319 * , 1025 ( 1997 ) .
gangadhara , r.t . ,
_ a&a _ * 314 * , 853 ( 1996 ) .
gangadhara , r.t . , and lesch , h. , _ a&a _ * 323 * , l45 ( 1997 ) .
lesch , h. , and birk , g.t . ,
_ a&a _ * 324 * , 461 ( 1997 ) .
machabeli , g.z . , and rogava , a.d .
, _ phys . rev .
a _ * 50 * , 98 ( 1994 ) .
melrose , d.b . , in : kirk , j.g . ,
melrose , d.b . , and priest , e.r .
, _ plasma astrophysics _ , edited by a.o .
benz , and t.j .- l .
couvoisier , springer , berlin , 1994 , pp .
113 - 223 .
rieger , f.m . , and mannheim , k. , _ a&a _ * 353 * , 473 ( 2000 ) . | centrifugal acceleration of charged test particles at the base of a rotating jet magnetosphere is considered .
based on an analysis of forces we derive the equation for the radial accelerated motion and present an analytical solution . it is shown that for particles moving outwards along rotating magnetic field lines , the energy gain is in particular limited by the breakdown of the bead - on - the - wire approximation which occurs in the vicinity of the light cylinder @xmath0 .
the corresponding upper limit for the maximum lorentz factor @xmath1 for electrons scales @xmath2 , with @xmath3 the magnetic field strength at @xmath0 , and is at most of the order of a @xmath4 for the conditions regarded to be typical for bl lac objects .
such values suggest that this mechanism may provide pre - accelerated seed particles which are required for efficient fermi - type particle acceleration at larger scales in radio jets . |
ultrashort optical pulses generate the coherent oscillation of the lattice , which modulates the macroscopic electric susceptibility .
these lattice oscillations are referred to as coherent phonons and can be detected with another ultrashort pulse via intensity modulations in reflectivity or transmissivity @xcite . using coherent phonons and a pump - probe type optical measurement
, we can directly observe the oscillation of the phonons and measure their dynamics for a wide variety of materials such as semimetals @xcite , semiconductors @xcite , superconductors @xcite and topological insulators @xcite .
in addition , the coherent phonons in carbon materials , e.g. , graphite @xcite , graphene @xcite , and carbon nanotubes @xcite , have attracted much attention in studies of electron - phonon coupling . to excite and measure the coherent phonons , the pulse duration of the pump and probe pulse needs to be shorter than the vibrational period of the phonons .
this requirement corresponds to a spectral width of the optical pulse that is much bigger than the phonon energy .
the coherent phonon dynamics can be observed as a change in transient reflectivity .
while this change depends on the probe light frequency , the reflected light can be measured without needing spectrally resolved methods .
this is commonly used and is referred to here as the spectrally integrated detection .
in contrast , the spectrally resolved detection shows that the associated change in transient reflectivity at shorter and longer wavelengths oscillates @xmath1 out of phase @xcite .
this implies that the sensitivity of the spectrally resolved detection is much higher than that of the spectrally integrated detection because the shorter- and longer- wavelength components cancel each out . in this paper , we investigate enhancement of the detection sensitivity for the coherent optical phonons using the spectrally resolved detection and found a strong enhancement of approximately @xmath0 times in diamond .
diamond has a wide band gap ( a direct gap of @xmath2 ev and an indirect gap of @xmath3 ev ) and a high optical - phonon frequency ( @xmath4 thz ) .
raman spectroscopy has been widely used to extract properties of diamond because the phonon spectrum is sensitively dependent on the crystal structure of carbon materials such as diamond , graphite , graphene , and carbon nanotubes . recently , as an application of quantum memory , the dynamics of longitudinal optical ( lo ) phonons in diamond has been studied @xcite . on the other hand ,
the coherent - phonon measurement using ultrashort visible pulses ( photon energy of @xmath5 ev ) and spectrally integrated detection have recorded @xmath4-thz coherent optical phonons in diamond and evaluated these lifetimes based on the difference in impurity rates @xcite . in the present experiment ,
we used an ultrashort infrared pulse with a central wavelength of @xmath6 ev and neglected linear and multiphoton absorption effects .
the coherent optical phonons are investigated using a pump - probe - type transient reflection measurement ( fig . [ expfig ] ) . the output pulse from the ti : sapphire oscillator ( femtolasers : rainbow ) , the spectrum of which was is given in fig .
[ spectrumexp ] measured using a usb spectrometer ( oceanoptics : usb2000 + ) , was divided into two pulses by a @xmath7 beam splitter , and used as pump and probe pulses .
the pump pulse went through a scan delay unit ( ape : scan delay 50 ) to control the time delay between the pump and probe pulses .
the scan delay was run with a sine function of @xmath8 hz .
also , the probe pulse was picked up by a @xmath9 beam splitter to measure the reference beam intensity at a photodiode ( pd1 ) .
thereafter , both pump and probe pulses were focused on the sample by using an off - axis parabolic mirror with a focal length of @xmath10 mm .
the reflected pulse from the sample was detected with a photodiode ( pd2 ) .
in addition , optical bandpass filters are put before pd2 for the spectrally resolved detection , which were fb740 , fb800 , fb850 , and fb900 ( thorlabs inc . ) with transmission at a central wavelength of @xmath11 , @xmath12 , @xmath13 , and @xmath14 nm , respectively , with a band width of @xmath15 nm . by applying the opposite bias voltages to pd1 and pd2 , we set the balanced detection before the experiment .
its differential signal , to be amplified with a low - noise current amplifier ( stanford research systems : sr570 ) , was measured by a digital oscilloscope ( iwatsu : ds5534 ) . to reduce the statistical error ,
the @xmath16 signals were averaged and taken as the measured value . by converting the temporal motion of the scan delay unit to the pump - probe pulse duration , the temporal evolution of the reflectivity change @xmath17 was obtained .
the time interval of the sampling data points was estimated to be @xmath18 fs .
note that the spectral chirping by the optics was compensated using a pair of chirp mirrors in order to minimize the pulse width at the sample position .
the ultrafast laser conditions in the following experiment were the spectral centroid of @xmath19 nm evaluated from fig .
[ spectrumexp ] , the pulse width @xmath20 fs from the frequency resolved auto correlation measurement ( femtolasers : femtometer ) , and the powers @xmath8 mw and @xmath21 mw of the pump and probe pulses , respectively . the power and polarization of both the pump and probe pulses were controlled using a half - wave plate and a polarizer .
the sample was a commercially available type - ib diamond crystal of a @xmath22 $ ] crystal plane ( sumitomo co. ) with a rectangular parallelepiped shape of face size with a @xmath23 mm @xmath24 @xmath23 mm , and a thickness of @xmath25 mm .
the polarization of the pump pulse was set parallel to the @xmath26 $ ] axis .
the polarizations of the pump and probe pulses are orthogonal to each other .
-fs laser pulse .
after the experimental procedures , this spectrum was measured at the output port of the ti : sapphire oscillator with a fiber - type spectrometer.,width=226 ]
we next analyze the coherent lo phonon dynamics by measuring the change in transient reflectivity .
figure [ typ ] shows the transient - reflectivity dynamics of the diamond obtained for a whole spectral range without any filters . with overlapping the pump and probe pulses there is a strong peak at zero delay because of diamond s nonlinear response . after the strong peaks
there is a modulation caused by the coherent optical phonons in diamond .
the transient reflectivity expanded over the interval from @xmath27 to @xmath12 fs clearly indicates that the vibrational period is @xmath28 fs ( frequency of @xmath29 thz ) .
the lifetime of this coherent oscillation was estimated to be @xmath30 ps by fitting the data with the damping oscillation .
this agrees well with a previous result ( @xmath31 ps ) @xcite .
fs and @xmath12 fs , and ( c ) the fourier spectrum of the data ( a ) in a range between 100 fs and 2400fs.,width=9 ] figure [ filters ] shows the spectrally resolved transient - reflectivity change @xmath32 with bandpass filters fb720 ,
fb800 , fb850 , and fb900 .
@xmath33 is the reflectivity measured with a filter without the pump pulse .
the amplitude of the coherent oscillations in the transient reflectivity obtained with the bandpass filter is larger than that without the band - pass filter , although the light intensity obtained with the band pass filter is reduced .
according to the estimation of the current intensity at pd2 , the ratio of the reflected probe pulse intensity with bf900 , bf850 , bf800 , and bf740 compared with that without filtering is estimated to be @xmath34 , @xmath35 , @xmath36 , and @xmath34 , respectively .
we demonstrate that the oscillation amplitude in @xmath32 with fb900 is @xmath0 times larger than that without the filter @xmath37 . of the type ib diamond obtained without any filters ( black ) and with filters fb900 ( red ) , fb850 ( orange ) , fb800 ( green ) , and fb740 ( blue ) .
the data are plotted with an offset.,width=8 ] the phase of the oscillation on the longer - wavelength side ( @xmath13 nm and @xmath14 nm ) is almost opposite to that on the shorter - wavelength side ( @xmath11 nm and @xmath12 nm ) .
this phase change in the transient transmission spectrum of carbon nanotubes was already reported @xcite .
the phase of the coherent phonon oscillation in the differential transmission for the radial breathing mode of single - walled carbon nanotubes depends on the detection wavelength .
the oscillations at wavelengths of @xmath38 nm and @xmath39 nm are anti - phase to each other , and hence below and above the resonance of the @xmath40 band gap . in contrast , in the present study , we observed a phase difference in the transient reflectivity although all the probe wavelengths are below the band gap . to elaborate a mechanism for our observed anti - phase effect , when a delay @xmath41 between the pump and probe pulses is much longer than the optical pulse width @xmath42 , the generation and the detection processes
are treated as independent . to clarify our observation in the spectrally resolved measurement ,
the detection process of the coherent phonons is only discussed .
the incidence of the pump pulse on the diamond at time @xmath43 induces coherent optical phonons via impulsive stimulated raman scattering @xcite because the optical energy ( @xmath6 ev ) is well below the band gap ( @xmath2 ev ) @xcite of the diamond .
the induced phonons modulate the polarization @xmath44 due to the probe pulse @xmath45 irradiated at @xmath46 as @xmath47 . here
, @xmath48 is proportioainal to the phonon coordinate , @xmath49 .
it is shown @xcite that the spectrally resolved reflection modulation @xmath50 is given by @xmath51 in which @xmath52 and @xmath53 are the fourier components of @xmath45 and @xmath48 , and @xmath54 is the central frequency of the probe pulse .
equation ( [ dr ] ) suggests that in @xmath50 , the optical frequency @xmath55 and the phonon frequency @xmath56 are synthesized into components of frequencies @xmath57 .
a quantum mechanical theory yields the formula for @xmath50 as @xmath58 e_0(\omega)\nonumber\\ & \times & \sin\omega\tau_p , \label{final1}\end{aligned}\ ] ] where @xmath59 is the dimensionless electron - phonon coupling constant , @xmath60 is the electric susceptibility of the crystal , and @xmath61 is a real quantity in the fourier transform limit . see technical details in appendix [ appendixa ] . in the present experiment ,
the peak position of the spectrum is @xmath12 nm ( @xmath62 thz ) , which is slightly smaller than the spectral centroid ( @xmath19 nm ) , and the frequency of the optical phonons is @xmath4 thz .
then , eq . ( [ final1 ] ) implies that the largest signals should be measured at @xmath63 thz and @xmath64 thz , corresponding to @xmath65 nm and @xmath66 nm , respectively , which is consistent with the present data .
it is worth noting that eq .
( [ final1 ] ) is a universal formula for the spectrally resolved detection of coherent phonons in the transparent region .
note that the prerequisite for the @xmath1 out - of - phase effect is that the electric susceptibility @xmath60 is a real quantity in the transparent region . in the opaque region ,
the relative phase of the oscillation depends on the relative magnitude of the real and imaginary parts of @xmath60 .
this formula clearly indicates that the change of the reflectivity at frequency @xmath57 oscillates @xmath1 out of phase as a function of pump - probe delay @xmath41 .
it is noted that ref .
@xcite gives a theoretical treatment of the pump - probe signal in the transmission process for optically thin systems .
essentially , the same formula can be applied to the reflection process by regarding @xmath67 as the electric field of the reflected probe pulse without the pump . in
the commonly used transient - reflectivity measurements without filters , the oscillations at higher and lower frequencies cancel out because they oscillate @xmath1 out of phase . while one may think that no coherent phonon oscillation is observed without any filters , the oscillation amplitude is proportional to the difference between the electric susceptibilities , @xmath68 , and the electric susceptibility
is an increasing monotonic function in the transparent region encountered in eq .
( [ final1 ] ) . in our case
, the lower frequency components can be observed in the detection without a filter . the oscillation amplitude in @xmath69 with the specific filter is bigger than that in @xmath37 without a filter .
our experimental data show that the sensitivity of the fb900 band - pass filter is @xmath0 times higher than that without any filter .
we remark that the observed oscillations with fb900 and fb850 are not perfectly @xmath1 out of phase .
this may be caused by the spectral chirping in the optical pulse because the fourier - transform - limited pulse was assumed for simplicity in our theoretical treatment .
we observed the changes in the transient reflectivity via coherent optical phonons in diamond using ultrashort infrared pulses and showed that the spectrally resolved detection can enhance the sensitivity of the coherent phonon measurement approximately @xmath0 times compared with the commonly used spectrally integrated measurement .
this is because the reflected - light intensities for higher and lower frequencies from the spectral centroid oscillate @xmath1 out of phase and canceled each other out . with our enhanced method
, we can measure the transient vibrational - state dynamics in a diamond more precisely , for example , the coherence time of the coherent phonon . as alluded to before @xcite ,
the vibrational state of a diamond is a candidate for room - temperature based quantum memory . by utilizing coherent optical phonons in diamond , a coherent read - write process in quantum memory
might be demonstrated using a pump - pump - probe method with wave - packet interference of optical phonons @xcite .
the authors thank yasuaki okano for providing useful discussions and also thank masaki hada and takuya imaizumi for experimental helps .
the authors thank kyoko kamo for help in the illustration of fig .
[ expfig ] .
this work was partially supported by core research for evolutional science and technology ( crest ) of the japan science and technology agency ( jst ) , jsps kakenhi grant numbers 25400330 , 14j11318 , 15k13377 , and 16k05410 , the collaborative research project of materials and structures laboratory , the daiko foundation , and the joint studies program of institute for molecular science , national institutes of natural sciences .
also , n. k. is partially supported by the japan society for the promotion of science .
the whole process of generation and detection of the coherent phonons may be formulated as a series of higher - order optical processes . in this appendix , we concentrate on the detection process , assuming that there exists a coherent lattice oscillation generated by a pump pulse irradiated in the past .
the coherent phonon is regarded as a tool to apply an oscillating perturbation on the electrons in the crystal , and the probe pulse detects the change in the electronic states through the modulation of the frequency dispersed signal .
it is noted that another quantum - mechanical treatment on the generation and detection mechanism of the coherent phonon was recently analyzed @xcite .
let us consider a two - band system for the insulating crystal , say diamond , which describes the dynamics in the probe process .
it is assumed that the energy of the excited states is modulated by the coherent oscillation of the lo phonon at the @xmath70-point , which has been generated by the pump pulse .
the model hamiltonian is given by @xmath71 where the state vector @xmath72 means the electronic ground state of the crystal , and @xmath73 is the state for which an electron with wave vector @xmath74 is excited from @xmath72 to the conduction band with the excitation energy @xmath75 .
the hamiltonian @xmath76 and @xmath77 are the phonon hamiltonians in the subspaces @xmath72 and @xmath73 . the creation and the annihilation operators of the lo phonon at the @xmath70-point with energy @xmath78
are denoted by @xmath79 and @xmath80 , respectively .
a similar model has been used to generate phononic states in quantum dots @xcite .
it is assumed that the excited states are coupled with the lo phonon mode through the deformation potential interaction with the dimensionless coupling constant @xmath81 . in approximation , we have neglected the @xmath74-dependence of the coupling constant , assuming a rigid - band shift because of the deformation potential interaction . in the bulk crystal , the huang
rhys factor @xmath82 is considered to be small , @xmath83 . within the rotating wave approximation , the interaction hamiltonian with the probe pulse
is given by @xmath84 in which @xmath85 is the transition dipole moment from @xmath72 to @xmath73 , and @xmath86 is the temporal profile of the electric field of the probe pulse with the delay time @xmath41 . the spectral profile of the probe pulse is given by @xmath87 with @xmath88 the interaction hamiltonian ( [ int ] ) can also be used also for the pump process by setting @xmath89 .
the only difference is that the amplitude of @xmath67 is much larger than that in the probe pulse . if one neglects the reflection from the back surface of the crystal , the pump - induced change of the reflection is obtained as a heterodyne modulation on the reflected probe pulse . aside from irrelevant factors ,
the spectrally resolved signal of the reflection @xcite is given by @xmath90 where @xmath55 is the detection frequency and @xmath53 is the fourier component of the modulated part of the polarization due to the coherent phonon , @xmath91 equation ( [ deltar ] ) can be derived from the argument that the small change in the reflected amplitude of the probe pulse originates from the loss or gain done by the work of the induced polarization in the presence of coherent lattice oscillation .
it can be shown that the spectrally integrated signal , which is obtained in the common pump - probe experiment for the coherent phonons , is given by the integration of the spectrally resolved signal as @xmath92 the actually observed change in the reflection is normalized by the reflection amplitude @xmath93 without the pump pulse . as @xmath94 , we find @xmath95 for non - resonant excitation by the pump pulse , the coherent oscillation of the optical phonons is driven by the impulsive momentum generated by the virtual transition to the excited electronic states .
we take the origin of time at the moment that the displacement of the phonon coordinate becomes maximum .
furthermore , we assume that the initial configuration of the phonon is given by the coherent state defined as @xmath96 for a real eigenvalue @xmath97 . while this is an approximation , it makes the whole calculation transparent .
hence , the pump pulse prepares the initial state in which the electron is in the ground state while the phonon is in the excited state . for this initial state wave function
@xmath98 the schrdinger equation @xmath99 is solved to the lowest order in @xmath100 as the third - order nonlinear polarization @xmath48 comes from the cross term of the first and second terms of eq .
( [ psi ] ) .
the time evolution and the expectation value of the phonon variable can be calculated by elementary arithmetic using the coherent - state representation .
we obtain @xmath106 \quad ( t \geq 0),\nonumber\\ & = & 0\quad ( t<0 ) , \label{p}\end{aligned}\ ] ] with a small positive number @xmath107 corresponding to the life time in the excited state .
we are not concerned with the relaxation mechanism of the coherent phonon in the present work .
the frequency - resolved polarization is given by eq .
( [ pomega ] ) . for large enough
@xmath41 , @xmath53 is written as @xmath108,\end{aligned}\ ] ] where we set @xmath109 and the optical response function @xmath110 the fourier transform of @xmath111 is the electric susceptibility of the material , @xmath112 applying the formula @xmath113 with @xmath114 being the @xmath115th bessel function , we find the final result in a closed expression , @xmath116\tau_p } e_0(\omega+m\omega - n\omega).\end{aligned}\ ] ] in bulk crystals , we may set @xmath117 and @xmath118 . in this case
, we retain only terms @xmath119 , @xmath120 and @xmath121 . as a function of @xmath41 ,
the oscillating part are given by terms @xmath122 and @xmath123 .
as @xmath124 , @xmath125 and @xmath126 , we find @xmath127.\end{aligned}\ ] ] because the electric susceptibility is a real quantity in the transparent region , @xmath128.\end{aligned}\ ] ] if the probe pulse is in the fourier transform limit , @xmath129 is also a real quantity and the above formula reduces to a much simpler expression , @xmath130 e_0(\omega)\sin\omega\tau_p . \label{final}\end{aligned}\ ] ] for a frequency at the peak position of the incident pulse @xmath54 , the peak frequency in the reflected signal becomes @xmath134 in the first term in eq .
( [ final ] ) , and @xmath135 in the second term .
the phase of the oscillation is reversed in both the low - frequency and the high - frequency parts .
furthermore , the oscillation amplitude at the high - frequency peak is slightly larger than the lower one if the pulse is in the fourier transform limit because the second derivative of the susceptibility is usually positive in the transparent region ; @xmath136 so that @xmath137 therefore , the change in transient reflectivity is not completely canceled out and can be measured in the commonly used spectrally integrated measurement .
99 t. dekorsy , g. c. cho , and h. kurz , in _ light scattering in solids viii _ , ( eds . ) m. cardona and g. gntherodt , ( springer , berlin , 2000 ) pp .
yan , e. b. gamble , and k. nelson , j. chem
. phys . * 83 * , 5391 ( 1985 ) .
t. k. cheng , j. vidal , h. j. zeiger , g. dresselhaus , m. s. dresselhaus , and e. p. ippen , appl .
lett . * 59 * , 1923 ( 1991 ) .
h. j. zeiger , j. vidal , t.k .
cheng , e. p. ippen , g. dresselhaus and m.s .
dresselhaus , phys .
b * 45 * , 768 ( 1992 ) .
m. f. decamp , d. a. reis , p. h. bucksbaum , and r. merlin , phys .
b * 64 * , 092301 ( 2001 ) .
h. katsuki , j. c. delagnes , k. hosaka , k. ishioka , h. chiba , e. s. zijlstra , m. e. garcia , h. takahashi , k. watanabe , m. kitajima , y. matsumoto , k. g. nakamura , and k. ohmori , nat .
commun . * 4 * , 2801 ( 2013 ) .
g. c. cho , w. ktt , and h. kurz , phys .
lett . * 65 * , 764 ( 1990 ) .
t. dekorsy , t. pfeifer , w. ktt , and h. kurz , phys .
b * 47 * , 3842 ( 1993 ) .
g. a. garrett , t. f. albrecht , j. f. whitaker and r. merlin , phys .
lett . , * 77 * , 3661 ( 1996 )
. o. v. misochko , k. kisoda , k. sakai , and s. nakashima , phys .
b * 61 * , 4305 ( 2000 ) .
m. hase , m. kitajima , a. m. constantinescu , and h. petek , nature * 426 * , 51 ( 2003 ) .
s. hayashi , k. kato , k. norimatsu , m. hada , y. kayanuma , and k. g. nakamura , sci .
rep . * 4 * , 4456 ( 2014 ) .
j. m. chwalek , c. uher , j. f. whitaker , g. a. mourou , and j. a. agostinelli , appl .
* 58 * , 980 ( 1991 ) .
w. albrecht , th .
kruse , and h. kurz , phys .
69 * , 1451 ( 1992 ) .
h. takahashi , y. kamihara , h. koguchi , t. atou , h. hosono , i. katayama , j. takeda , m. kitajima , and k. g. nakamura , j. phys .
80 * , 013707 ( 2011 ) .
a. q. wu , x. xu , and r. venkatasubramanian , appl .
. lett . * 92 * , 011108 ( 2008 ) .
n. kamaraju , s. kumar , and a. k. sood , europhys . lett .
* 92 * , 47007 ( 2010 ) . k. norimatsu , j. hu , a. goto , k. igarashi , t. sasagawa , and k. g. nakamura , solid state commun .
* 157 * , 58 ( 2013 ) .
o. v. misochko , j. flock , and t. dekorsy , phys .
b * 91 * , 174303 ( 2015 )
. k. ishioka , m. hase , m. kitajima , l. wirtz , a. rubio , and h. petek , phys .
b * 77 * , 121402(r ) ( 2008 ) .
i. katayama , k. sato , s. koga , j. takeda , s. hishita , h. fukidome , m. suemitsu , and m. kitajima , phys .
b , * 88 * , 245406 ( 2013 ) .
kim , a.r.t .
nugraha , l.g .
booshehri , e.h .
hroz , k. sato , g. d. sanders , k .- j .
yee , y .- s .
lim , c. j. stanton , r. saito , and j. kono , chem .
phys . * 413 * , 55 ( 2013 ) .
k. kato , k. ishioka , m. kitajima , j. tang , r. saito , and h. petek , nano lett . * 8 * , 3102 ( 2008 ) .
a. gambetta , c. manzoni , e. menna , m. meneghetti , g. cerullo , g. lanzani , s. tretiak , a. piryatinski , a. saxena , r.l .
martin , and a. r. bishop , nat .
* 2 * , 515 ( 2006 ) .
g. d. sanders , c. j. stanton , j .- h .
kim , k .- j .
yee , y .- s .
lim , e. h. hroz , l. g. booshehri , j. kono , and r. saito , phys .
b * 79 * , 205434 ( 2009 ) .
k. makino , a. hirano , k. shiraki , y. maeda , and m. hase , phys .
b * 80 * , 245428 ( 2009 ) . j
kim , k .- j .
kim , k .- j .
yee , y .- s .
lim , g. d. sanders , c. j. stanton , l. g. booshehri , e. h. hroz , and j. kono , phys .
* 102 * , 037402 ( 2009 ) .
r. merlin , solid state commun . * 102 * , 207 ( 1997 ) .
o. v. misochoko , m. hase , and m. kitajima , j. phys . : condens .
matter * 16 * , 1879 ( 2004 ) .
k. mizoguchi , r. morishita , and g. oohata , phys .
lett . * 110 * , 077402 ( 2013 ) . k. c. lee , b. j. sussman , j. nunn , v. o. lorenz , k. reim , d. jaksch , i. a. walmsley , p. spizzirri , and s. prawer , diam .
mater , * 19 * , 1289 ( 2010 ) .
k. c. lee , m. r. sprague , b. j. sussman , j. nunn , n. k. langford , x .-
jin , t. champion , p. michelberger , k. f. reim , d. england , d. jaksch , i. a. walmsley , science * 334 * , 1253 ( 2011 ) .
k. c. lee , b. j. sussman , m. r. sprague , p. michelberger , k. f. reim , j. nunn , n. k. langford , p. j. bustard , d. jaksch and i. a. walmsley , nat .
* 6 * , 41 ( 2012 ) .
d. g. england , p. j. bustard , j. nunn , r. lausten , and b. j. sussman , phys . rev .
111 * , 243601 ( 2013 ) .
d. g. england , k. a. g. fisher , j .-
w. maclean , p. j. bustard , r. lausten , k. j. resch , and b. j. sussman , phys .
114 * , 053602 ( 2015 ) .
k. ishioka , m. hase , m. kitajima , and h. petek , appl .
. lett . * 89 * , 231916 ( 2006 ) .
k. g. nakamura , y. shikano , and y. kayanuma , phys .
b * 92 * , 144304 ( 2015 ) .
r. p. milden , in _ optical engineering of diamond _ , ( eds . ) r. p. milden and j. r. rabeau ( wiley - vch verlag & co. kgaa , weinheim , 2013 ) , pp . 134
. g. stock and w. domcke , phys .
a * 45 * , 3032 ( 1992 ) .
s. j. rosenthal , x. xie , m. du , and g. r. fleming , j. chem .
phys . * 95 * , 4715 ( 1991 ) .
k. ohmori , annu .
phys . chem . *
60 * , 487 ( 2009 ) .
m. esposito , k. titimbo , k. zimmermann , f. giusti , f. randi , d. boschetto , f. parmigiani , r. floreanini , f. benatti , d. fausti , nature comm .
* 6 * , 10249 ( 2015 ) .
d. e. reiter , d. wigger , v. m. axt , and t. kuhn , phys .
b * 84 * , 195327 ( 2011 ) . | coherent optical phonons in bulk solid system play a crucial role in understanding and designing light - matter interactions and can be detected by the transient - reflectivity measurement . in this paper , we demonstrate spectrally resolved detection of coherent optical phonons in diamond from ultrashort infrared pump - probe measurements using optical band - pass filters .
we show that this enhances the sensitivity approximately @xmath0 times in measuring the coherent oscillations in the transient reflectivity compared with the commonly used spectrally integrated measurement .
to explain this observation , we discuss its mechanism . |
in the standard cold dark matter ( cdm ) paradigm of structure formation , galaxies are supposed to form and evolve in dark matter halos .
the study of the clustering properties of dark matter halos and their relation to galaxy clustering can thus help us to understand the connection between halos and galaxies , and hence to understand how galaxies form and evolve in dark matter halos .
it is now well known that the correlation strength of dark matter halos depends strongly on halo mass ( e.g. , mo & white 1996 ; mo et al .
1997 ; jing 1998 ; sheth & tormen 1998 ; sheth , mo & tormen 2001 ; seljak & warren 2004 ) , and this dependence , which is referred to as the halo bias , has been widely used to understand the clustering of galaxies via the halo occupation model ( e.g. , jing , mo & brner 1998 ; peacock & smith 2000 ) , and the conditional luminosity function model ( e.g. , yang , mo & van den bosch 2003 ) .
more recently , a number of independent investigations have shown that the halo bias depends not only on the mass but also assembly time of dark matter halos , in the sense that halos of a given mass , particularly low - mass ones , are more strongly correlated if they assembled half of their masses earlier ( e.g. gao et al .
2005 ; harker et al .
2006 ; zhu et al .
2006 ; wechsler et al . 2006
; jing , suto & mo 2007 ; wetzel et al .
2007 ; bett et al . 2007
; gao et al . 2007 ; li et al . 2008 ) .
the origin of this assembly - time dependence of the halo bias , referred to in the literature as the halo assembly bias , is important to understand , because if it is due to a process that can also affect galaxy formation and evolution in a halo then it would affect our interpretation of galaxy clustering in terms of halo clustering .
there have been theoretical investigations about the origin of the halo assembly bias ( e.g. wang et al .
2007 ; sandvik et al . 2007 ; desjacques 2007 ; keselman & nusser 2007 ; dalal et al .
2008 ; hahn et al . 2008 ) .
wang et al .
( 2007 ) find that old , small halos tend to live near massive halos , and suggest that the tidal truncation of accretion may be responsible for the assembly bias .
this is consistent with the result of maulbetsch et al .
( 2007 ) , who find that most of the small halos in high - density regions have ceased accretion . along this line ,
desjacques ( 2007 ) develops an ellipsoidal collapse model which takes into account the large - scale tidal field , while keselman & nusser ( 2007 ) perform simulations using the zeldovich ( pz ) approximation to take into account the large - scale tidal effects .
both investigations find significant dependence of halo bias on halo assembly history , indicating that large - scale tidal effects may play an important role in producing the assembly bias .
more recently , ludlow et al . ( 2008 , hereafter l08 ) study in detail 5 simulations of dark matter halos and find that a significant fraction of small halos are physically associated with nearby massive halos .
these small halos have been once inside their host halos but were ejected due to interactions with other subhalos ( see also lin et al .
2003 ; wang et al . 2005
; gill , knebe , & gibson 2005 ) .
l08 suggest that these ejected subhalos may be responsible for the enhanced clustering of old small halos .
however , because of the small volume of their simulations , they were not able to quantify whether the ejected population alone can account for the assembly bias seen in cosmological simulations .
in this paper we use a high - resolution @xmath0-body simulation in a cosmological box to study the distribution of ejected subhalos in space and to quantify the contribution of this population of halos to the assembly bias .
the outline of the paper is as follows . in section [ sec_sim ]
we describe briefly the simulation used and how ejected subhalos are identified . in section [ sec_rel ]
we study the distribution of the ejected subhalos in phase space , and how the distribution depends on the properties of their hosts . in section [ sec_ori ]
we examine the contribution of the ejected subhalos to the assembly bias obtained in our simulation .
finally , in section [ sec_dis ] , we discuss and summarize our results .
the simulation used in this paper is obtained with the @xmath4 code described in jing & suto ( 2002 ) .
it assumes a spatially - flat @xmath5cdm model , with density parameters @xmath6 and @xmath7 .
the cdm power spectrum is assumed to be that given by bardeen et al .
( 1986 ) with a shape parameter @xmath8 and an amplitude specified by @xmath9 .
the cdm density field is traced with @xmath10 particles , each having a mass @xmath11@xmath12 , in a cubic box of 100 @xmath13 .
the softening length is @xmath14 ( s2 type ) .
the simulation , started at redshift 72 , is evolved with 5000 time steps to the present day ( @xmath15 ) and has 60 outputs from @xmath16 , equally spaced in @xmath17 .
dark matter halos were identified with a friends - of - friends algorithm with a link length that is @xmath18 times the mean particle separation .
our analysis focuses on the ejected subhalos which are identified as fof halos at redshift @xmath1 . in order to determine whether a halo is an ejected subhalo or a normal halo , a detailed merging tree for each fof halo
is required so that we can trace a fof halo back in time to see whether it has ever been within another fof halo .
we consider a halo at any given redshift @xmath19 to be a progenitor of a descendant halo in the next output , if more than half of its particles are found in the descendant . a given halo in general
can have one or more progenitors but only one descendant .
we can therefore use the uniqueness of the descendant to build up the merging tree for each halo .
each fof halo at the present time has one and only one merging tree to describe its assembly history .
there is a small fraction of halos at @xmath19 that have dispersed and have no descendant at @xmath1 .
these halos are excluded in our analysis .
the merging tree of a halo can be used to verify whether an isolated fof halo ( called halo ` a ' ) at @xmath1 was accreted into a massive halo earlier . to do this , we search in the next snapshot ( in reverse order of time ) a ` host ' halo that contains at least half of the particles in halo ` a ' , but does not belong to the merging tree of halo ` a ' .
if no such a ` host ' halo is found in this snapshot , we take the most massive progenitor of halo ` a ' in this snapshot and repeat the same procedure for this progenitor as we have carried out for halo ` a ' .
this procedure is continued until a ` host ' halo is found or the tree ends .
if such a ` host ' halo is found , halo ` a ' is then identified as an ejected subhalo , which is said to be ejected at the time , @xmath20 , when the ` host ' halo is found .
note that the ejection time , @xmath20 , is defined to be the time of the snapshot at which the ` host ' and ` ejected ' halos were just separated from the same fof group .
we also define an accretion time , @xmath21 , as the time when half of the particles in the most massive progenitor of halo ` a ' at ejection time is first contained in the most massive progenitor of its ` host ' . if no ` host ' halo is found before the merging tree ends , halo ` a '
is said to be a normal halo . applying this method to all halos at @xmath1
, we construct a catalogue of ejected subhalos , and the total number of ejected subhalos is listed in table [ tab_nh ] in three mass ranges .
we denote the masses of an ejected subhalo and of the corresponding ` host ' halo at any time @xmath22 as @xmath23 and @xmath24 , respectively , and we use @xmath25 to denote the mass of a normal halo at redshift @xmath1 .
thus , @xmath26 is the mass of the ` host ' halo at the time of ejection , and @xmath27 is the mass of the descendant of the ` host ' at the present time , @xmath28 . in fig .
[ fig_r12 ] we show @xmath29 versus @xmath30 for halos with @xmath31 .
the corresponding histograms for @xmath30 and @xmath32 are also shown in the figure .
the results for other mass ranges are similar and are not shown . as one can see ,
the majority of the ejected subhalos were ejected by ` host ' halos that are much more massive than the ejected subhalos themselves .
only small fraction of halos are ejected by host halos with comparable masses : about 11% ( 34% ) of ejected subhalos have a mass larger than 0.5 ( 0.1 ) times of their host halo mass .
furthermore , most of the ejected halos have masses that are similar to those at the time of ejection , indicating that mass loss or accretion after ejection is not severe .
however , the distribution has two extended tails , and the 10th and 90th percentile values are 0.73 and 1.48 , respectively . in our analysis , we remove all systems with @xmath33 , i.e. the halos that have accreted more than their initial masses after ejection .
note that the removed halos is a very small fraction , about 5% , of the total sample ( table [ tab_nh ] ) . including them
does not have any significant impact on the results to be presented below .
after the removal , the final numbers of the ejected subhalos in different halo mass ranges are also given in table [ tab_nh ] .
for comparison , we also list the corresponding numbers for all ( ejected plus normal ) halos in these mass ranges . about 9% - 4% of all the halos in the mass ranges considered
are ejected subhalos , and the fraction decreases with increasing @xmath34 .
.the number of ejected subhalos in various samples [ cols="<,^,^,^",options="header " , ] it is possible that an ejected subhalo after ejection can exchange mass with nearby halos or with the background density field , so that some of the particles contained in the ejected subhalo at the present time may not be contained in its ` host ' halo . in order to quantify this
, we consider an ejected mass , @xmath35 , which is defined as the mass of particles which are contained both in the ejected subhalo at @xmath1 and in its ` host ' halo at the ejection redshift .
the ratio between @xmath35 and @xmath36 can then be used as a measure of the fraction of retained mass of an ejected subhalo . in fig .
[ fig_fm ] we show the probability distribution of @xmath37 .
as one can see , most of the ejected subhalos at @xmath1 can retain more than @xmath38 of their original masses at ejection .
this also shows that our method for identifying ejected subhalos is valid , in the sense that most of their masses were indeed once contained in their hosts .
we define the assembly redshift , @xmath39 , of a halo at redshift @xmath1 as the redshift when its most massive progenitor first reaches half of the final mass of the halo .
if necessary , interpolation between two adjacent outputs is used .
here we do not use the merging tree described above to search for the progenitor .
instead we use the method described in wang et al .
( 2007 ; see also hahn et al .
2008 ) . very briefly , a halo in any given output at @xmath19 is considered to be a progenitor of the halo at @xmath1 if more than half of its particles are found in the final halo . in most cases , the merging trees constructed with this method
is very similar to that constructed with the method described above .
the advantage of the method adopted here is that it can trace an ejected subhalo backwards to the time even before it is accreted by its host halo .
since ejected halos may be strongly stripped by their host halos and lose part of their masses , the assembly redshifts of the ejected subhalos are generally higher than than their ejection redshifts and can be identified with this method .
in fig . [ fig_nt ] , we show the histograms of the scaled time period an ejected subhalo stayed in its host : @xmath40 . here
@xmath41 is the dynamical time at @xmath42 of each host halo at the ejection time , with @xmath42 the virial radius within which the mean overdensity of the halois 200 times the critical density . the dynamical time is defined as : @xmath43 where @xmath44 is the circular velocity at @xmath42 and @xmath45 is the hubble constant , at the ejection time .
as one can see , the time period an ejected subhalo stays within its host ranges from @xmath46 to @xmath47 times the dynamical time . to examine this time period in more detail , we split the ejected subhalos sample into two subsamples according to the @xmath48 ratio .
the red ( dark ) and green ( light ) histograms in fig .
[ fig_nt ] represent the results with @xmath49 and @xmath50 , respectively .
evidently , the distribution for the case with @xmath51 is peaked around 2 , and only 24% of halos have @xmath52 . on the other hand , halos with large @xmath53 ratio show quite a different distribution .
it peaks at @xmath54 0.6 , and the fraction of halos with @xmath55 is more than 55% .
this indicates that many of the ejected subhalos with a large @xmath56 may be flybys which happen to be close to another halo and be linked to it by the fof group finder .
they may , therefore , represent a population that is different from the population with a small @xmath57 , which most likely have run through their hosts . in what follows , we will treat these two populations separately .
it should be pointed out , however , that there is an excess at 0.6 for halos with @xmath58 , and there is a long tail for the subsample with @xmath59 ( see fig .
[ fig_nt ] ) .
thus , the @xmath60 ratio alone can not distinguish the two populations unambiguously .
we show the probability distribution of ejected subhalos in their distances to the host halos at @xmath1 in fig .
[ fig_nrv ] . here , the distance is scaled by @xmath42 of each host halo at redshift @xmath1 .
we use the scaled distance @xmath61 instead of @xmath62 , because @xmath42 is the only important length scale related to the dynamics of a virialized halo . as
one can see , the distribution peaks at @xmath63 .
this is different from what is shown in l08 , because their result includes also subhalos within host halos .
most of the ejected subhalos are distributed within @xmath64 , in good agreement with the finding of l08 , but the distribution has a long tail at large distance .
our detailed examination shows that the long tail is dominated by ejected subhalos that have masses not much smaller than those of their ` hosts ' at ejection .
if we _ exclude _ the ejected subhalos with @xmath65 , then the extended tail disappears as long as @xmath66 is chosen to be smaller than @xmath67 .
the distribution obtained for @xmath68 , @xmath69 and @xmath70 are shown in fig.[fig_nrv ] for comparison .
this result indicates that the two populations of the ejected subhalos discussed above have different distribution .
ludlow et al.(2008 ) found that subhalos with small mass at accretion time is less centrally concentrated than their massive counterparts .
we do not find any significant evidence for such dependence .
the difference in the result may be due to the difference in the halo definition and , particularly , in the contamination of the flybys population found in our simulation . this flyby population may be not important in l08 because of their way of selecting halos ( see section 2.2 in their paper ) .
ejected subhalos in general are distributed around massive halos , but there are also normal halos which are distributed near massive halos but have never been inside a massive halo . in fig .
[ fig_frv ] , we show the average fraction of ejected subhalos around their host halos . this is the ratio between the number of ejected subhalos and that of the total population ( ejected plus normal ) as a function of the scaled distance to the host halos .
the left panel shows the average fractions of ejected subhalos with @xmath71 around host halos in three mass ranges . in the right panel
, we show the results for ejected subhalos with @xmath72 . in the mass ranges probed here ,
the ejected fraction as a function of the scaled distance is insensitive to the masses of both the ejected subhalos and the host halos .
the fraction is between 30% and 75% in the range @xmath73 and between 10% and 40% in the range @xmath74 .
these results are in agreement with those of l08 , who found fractions of about 65% and 33% in the similar distance ranges ( see also gill et al .
2005 ; wang et al . 2005 ) , although the mass ranges for both the host halo and the ejected subhalos considered by them are quite different from what we are considering here . note that , at large scaled distances , the fraction of ejected subhalos around host halos with the lowest mass is larger than that around massive host halos .
this is again due to the population in which the ejected subhalos have masses comparable to their hosts at the time of ejection . in fig .
[ fig_vrr ] we show the scaled radial velocity , @xmath75 , versus scaled distance for both ejected and normal halos within @xmath76 of the host halos of ejected halos . here
@xmath44 is the circular velocity of each host halo , and @xmath77 is the relative _ peculiar _ velocity along the separation between a host and an ejected or a normal halo .
we also split the ejected subhalos and normal halos into four subsamples according to the scaled distance and calculate the average velocity and average distance for each subsample .
the results are shown as the big symbols in the figure .
we consider two mass ranges of host halos , @xmath78 and @xmath79 .
note that some normal halos may be counted more than once since they may be within @xmath80 of more than one host halo .
clearly , the ejected subhalos and normal halos show different distribution in the phase space .
the radial velocity distribution of ejected subhalos associated with massive hosts is quite symmetric and quite independent of the distance to the hosts , suggesting that on average there is about equal chance for an ejected subhalo to be moving away from or falling back towards the host .
it is also consistent with the results of l08 although the velocities shown in their paper are the peculiar velocities plus the hubble flow .
the average radial velocity of ejected halos associated with low - mass hosts increase with the distance to the hosts .
this is because of the contamination of the flybys mentioned above .
the behavior of normal halos is quite different .
most of the normal halos are preferentially moving towards the host halos , as indicated by the systematically negative values of @xmath77 .
the average value of @xmath81 is about @xmath82 , independent on the distance and the mass of the hosts .
the difference here suggests that the ejected subhalos are a distinctive population among the total halo population .
a careful search shows that the scaled velocity dispersion of normal halos around small host halos is larger than that around massive host halos .
this may be due to the fact that environmental effects play a more important role around smaller hosts .
gao et al .
( 2004 ) find that the abundance of subhalos within host halos decreases with the formation redshift of the hosts , indicating that subhalos in early - formed host halos are more likely to be disrupted .
however , as pointed out by l08 , the subhalos within a host halo represent a rather incomplete census of the substructure physically related to the host .
it is thus important to check whether the abundance of the ejected subhalos is also related to the assembly time of their hosts . in fig .
[ fig_nm ] , we show the number of ejected subhalos as a function of the assembly redshift of the host halo . the left panel is the result for host halos with @xmath83 and @xmath84 . as one can see
, there is a clear trend that older host halos tend to eject more subhalos , although the scatter is quite large . to compare with gao et al .
( 2004 ) , we also show the results with @xmath85 in the right panel of fig.[fig_nm ] .
the trend is weaker , likely due to the statistical variation caused by small numbers .
this trend suggests that the old host halos tend not only to destroy their subhalos , but also to eject them .
the decrease of subhalo abundance with formation redshift observed by gao et al .
( 2004 ) is therefore a result of both subhalo ejection and destruction . for each host
, the number of destroyed subhalos , @xmath86 , reads , @xmath87 where @xmath88 is the number of total accreted subhalos , and @xmath89 and @xmath90 are the numbers of surviving subhalos within the host and of ejected subhalos , respectively .
thus , in order to quantify the importance of ejection and destruction , one needs detailed merging trees to trace the evolution of subhalos within their hosts .
due to the strong interaction with their hosts before ejection , ejected subhalos may lose part of their mass .
since it is difficult for them to accrete more mass after ejection , the ejected subhalos should have acquired most of their mass before the ejection .
thus , on average these halos should have assembly times ( defined in section [ sec_ass ] ) that are earlier than their normal counterparts . in order to show this ,
we first split the halo sample in a certain mass range into 10 equal - sized subsamples according to assembly redshift .
we then calculate the fraction of ejected halos and the average assembly redshift for each of these subsamples and show the fraction versus the redshift in fig .
[ fig_fz ] . among the 10 percent halos with the highest assembly redshifts , about @xmath91 to @xmath92
are ejected subhalos depending on the subhalo mass in question .
the fraction decreases rapidly down to about @xmath93 - @xmath94 for halos with the lowest assembly redshifts .
since the ejected subhalos are expected to be located in high - density regions due to their associations with massive halos , they are expected to be strongly clustered .
the presence of this population of ejected subhalos may therefore contribute to the assembly bias seen in cosmological @xmath0-body simulations ( e.g. gao et al .
2005 ; jing , suto & mo 2007 ) . in order to examine whether or not the ejected subhalos are fully responsible for the assembly bias
, we estimate the halo bias as a function of assembly redshift separately for all halos , normal halos and ejected ( sub)halos .
we first estimate the mean overdensity of dark matter within a sphere of radius @xmath95 around each dark matter halo , @xmath96 , and then measure the bias parameter of a given set of halos using @xmath97 where @xmath98 is the average overdensity around the set of halos in question , and @xmath99 is the average overdensity within all spheres of radius @xmath95 centered on dark matter particles .
this method has been demonstrated to reproduce the bias obtained using auto - correlation function of halos ( wang et al .
the results are shown in fig.[fig_bz ] .
as expected , the ejected subhalos in general have a much higher bias parameter than other halos of similar masses .
the bias parameter for ejected subhalos with @xmath100 is lower ( blue dash dot lines ) , presumably because they are associated with hosts of lower masses , as mentioned above .
however , even if all ejected subhalos are excluded , the assembly bias is still significant ( see the red dash lines in fig .
[ fig_bz ] ) . for halos with masses larger than @xmath101 ,
the assembly bias is almost entirely due to normal halos instead of ejected subhalos . even for halos of lower masses ( e.g. @xmath102 ) ,
including ejected only increases the bias parameter by @xmath103 .
the reason is that the fraction of ejected subhalos is small among the total halo population of the same mass .
thus , ejected subhalos can not explain the full range of assembly bias seen in cosmological @xmath0-body simulations , even though they are strongly clustered in space .
this result is consistent with the finding of wang et al .
( 2007 ) that the assembly bias is mainly due to the fact that small old halos tend to live in the vicinity of massive halos and their growth at low redshift is suppressed by the large - scale tidal field .
an analysis along this line is performed by keselman & nusser ( 2007 ) .
they used the punctuated zeldovich approximation , in which the highly non - linear effects such as tidal stripping are excluded , to run simulations of structure formation , and found that the assembly bias is still present .
all these indicate that the large - scale environmental effects have also played an important role in the formation of some normal halos ( see also hahn et al .
in this paper , we use a high - resolution cosmological simulation to study the distribution of the ejected subhalos , their connection to the host halos , and their contribution to the halo assembly bias seen in cosmological simulations .
our main results are summarized as follows . 1 .
the fraction of the ejected subhalos in the total halo population of the same mass ranges from 9% to 4% for halo masses from @xmath2 to @xmath3 .
the time period an ejected subhalo stays in its host has wide distribution , ranging from less than 0.5 to more than 4 times the dynamical time of of the host .
the distribution peaks at 2 and 0.6 for ejected subhalos with @xmath58 and @xmath50 , respectively , indicating the existence of two distinctive populations of ejected halos .
most of the ejected subhalos are found to be distributed within about 4 times the virial radius of their hosts .
the fraction of ejected subhalos is about 30% - 75% within the distance range between @xmath42 and @xmath104 to the hosts and about 10% - 40% in the distance range @xmath74 .
the radial velocity distribution of ejected subhalos is quite symmetric , while normal halos of the same mass generally tend to fall onto nearby massive halos .
the number of subhalos ejected from a host of given mass increases with the assembly redshift of the host .
thus , subhalos tend to be less abundant in halos that formed earlier , not only because subhalos in such hosts are more likely to be destroyed , but also because they are more likely to be ejected . 6 . ejected subhalos in general
reside in high - density regions , and have a much higher bias parameter than normal halos of the same mass .
they also have earlier assembly times , so that they contribute to the assembly bias of dark matter halos seen in cosmological simulations .
the assembly bias is _ not _ dominated by the ejected population .
this indicates that large - scale environmental effects may also be important in the formation of normal halo population , and in producing the assembly bias .
the results obtained here may have important implications to the understanding of galaxy distribution in the cosmic density field . as a subhalo pass through a massive host , tidal and/or ram - pressure stripping by the host halo
may get rid of the gas reservoir in the ejected halo , thereby quenching star formation in it .
the situation may be similar to what happens to satellite galaxies , although the ejected galaxies are not observed as satellites in massive halos .
if the quenching processes are effective , we would expect a population of faint red galaxies that are not contained in any massive halos but are distributed around them .
it is therefore interesting to see if such a population of galaxies does exist . in a recent investigation ,
wang et al .
( 2008 ) found that the reddest @xmath105 - @xmath106 among all faintest galaxies are physically associated with massive halos .
about half of this population resides within massive halos as satellites .
the other half resides outside massive halos and are distributed within about 3 times the virial radii of their nearest massive halos .
very likely , this population of galaxies are hosted by the ejected subhalos we are studying here .
clearly , it is interesting to study further the connection between ejected subhalos and this population of galaxies , so as to understand how environmental effects operate on satellite galaxies in their host halos .
we thank the anonymous referee for his / her constructive comments .
hyw would like to acknowledge the support of the knowledge innovation program of the chinese academy of sciences , grant no .
kjcx2-yw - t05 and nsfc 10643004 .
hjm would like to acknowledge the support of nsf ast-0607535 , nasa aisr-126270 and nsf iis-0611948 .
ypj is supported by nsfc ( 10533030 , 10821302 , 10878001 ) , by the knowledge innovation program of cas ( no .
kjcx2-yw - t05 ) , and by 973 program ( n o.2007cb815402 ) . | using a high - resolution cosmological @xmath0-body simulation , we identify the ejected population of subhalos , which are halos at redshift @xmath1 but were once contained in more massive ` host ' halos at high redshifts .
the fraction of the ejected subhalos in the total halo population of the same mass ranges from 9% to 4% for halo masses from @xmath2 to @xmath3 .
most of the ejected subhalos are distributed within 4 times the virial radius of their hosts .
these ejected subhalos have distinct velocity distribution around their hosts in comparison to normal halos . the number of subhalos ejected from a host of given mass increases with the assembly redshift of the host .
ejected subhalos in general reside in high - density regions , and have a much higher bias parameter than normal halos of the same mass .
they also have earlier assembly times , so that they contribute to the assembly bias of dark matter halos seen in cosmological simulations . however , the assembly bias is _ not _ dominated by the ejected population , indicating that large - scale environmental effects on normal halos are the main source for the assembly bias .
[ firstpage ] dark matter - large - scale structure of the universe - galaxies : haloes - methods : statistical |
a merger between two galaxies , each with its own central supermassive black hole ( smbh ) , results in a merger - remnant galaxy hosting two smbhs . as these smbhs orbit in the potential of the host galaxy , they are dragged closer together by dynamical friction from the surrounding stars , form a gravitationally - bound binary system , and ultimately merge @xcite . while the smbhs are at kiloparsec ( kpc ) scale separations , before they are bound as a binary system , they are known as dual smbhs .
gas churned up by the galaxy merger may accrete onto the dual smbhs , and cases where one or both of the smbhs fuel active galactic nuclei ( agns ) are known as offset agns and dual agns , respectively .
these offset and dual agns , which we refer to here as having separations @xmath9 kpc , have valuable potential as new observational tools for studies of galaxy evolution , including measurements of galaxy merger rates , smbh mass growth , and smbh merger rates . however ,
because of their small spatial separations from the host galaxy centers , offset and dual agns are challenging to identify . as a consequence , early discoveries of offset agns and dual agns were serendipitous @xcite . while ultra - hard x - rays have been used to uncover a handful of dual agns @xcite , serendipitous discoveries of dual agn candidates continue today ( e.g. , @xcite ) . to accelerate the discovery rate of offset and dual agns , a new systematic approach was developed to identify candidates in spectroscopic surveys of galaxies .
this technique uses galaxy spectra to select offset and dual agn candidates as narrow agn emission lines with line - of - sight velocity offsets of a few hundred km s@xmath10 relative to the host galaxy stellar absorption features .
offset agn candidates display velocity offsets in single - peaked narrow agn emission lines relative to stellar absorption lines , while the dual agn candidates have double - peaked narrow lines .
numerical simulations of agn activity in galaxy mergers show that these double - peaked lines are indeed useful selectors of dual agns @xcite .
the velocity - offset approach was first applied systematically to the deep2 galaxy redshift survey , where it was used to identify 30 offset agn candidates @xcite and two dual agn candidates @xcite at @xmath11 .
subsequently , it was used in the sloan digital sky survey ( sdss ) to uncover 340 unique double - peaked agns at @xmath12 @xcite and 131 double - peaked quasars at @xmath13 @xcite .
these double - peaked agns have been a reservoir for follow - up studies aimed at distinguishing line profiles produced by dual agns from those produced by narrow - line region kinematics ( e.g. , @xcite ) .
some of these observations have resulted in confirmations of dual agns @xcite , but the majority of double - peaked agns are likely produced by gas kinematics associated with a single agn ( e.g. , @xcite ) . given the successes of using velocity - offset agn emission lines to select offset and dual agn candidates in deep2 and sdss , it is a natural extension to apply the same approach to other large spectroscopic surveys of galaxies , such as the agn and galaxy evolution survey ( ages ) . a search for offset and dual agn candidates in ages , at a mean redshift of @xmath14 , would fill the gap between the lower redshift sdss ( @xmath15 ) and higher redshift deep2 ( @xmath16 ) samples of dual smbh candidates , enabling dual smbhs to be used for studies of galaxy evolution across the full range of @xmath17 . here
we present the results of our systematic search for velocity - offset narrow agn emission lines in ages galaxy spectra , which uncovered five offset agn candidates and two dual agn candidates .
these galaxies , at @xmath18 , are promising candidates for follow - up observations that will definitively determine whether they host offset and dual agns .
the remainder of this paper is organized as follows . in section 2
, we describe the ages spectra , agn selection , and our measurements of the redshifts , emission line fluxes , and velocity offsets of the emission lines . in section 3
, we present our identifications of two double - peaked agns and five offset agn candidates , examine their host galaxies , and compare them to similar candidates in other large spectroscopic surveys of galaxies .
section 4 gives our conclusions .
we assume a hubble constant @xmath19 km s@xmath10 mpc@xmath10 , @xmath20 , and @xmath21 throughout , and all distances are given in physical ( not comoving ) units .
our sample consists of a catalog of optical galaxy spectra observed for ages @xcite . using hectospec ,
an optical fiber - fed spectrograph with @xmath22 fibers on the mmt 6.5 m telescope , ages observed 7.7 deg@xmath23 of the botes field in the noao deep wide - field survey @xcite .
the resultant spectra have a wavelength coverage of 3700 9200 , and the spectral resolution is 6 , yielding @xmath24 .
ages determined spectroscopic redshifts for 18,163 galaxies to a limiting magnitude of @xmath25 . since we will use the , , , and @xmath26emission lines to diagnose agn activity ( see section [ agn ] ) , we select the galaxy spectra where all four of these emission lines are within the ages wavelength range .
this cut results in 8136 spectra at @xmath27 , and this sample is the focus of our analysis as described below .
while redshifts for the ages galaxies have already been measured by cross correlation with emission and absorption line galaxy and agn template spectra @xcite , these redshifts may be weighted towards the emission lines and hence not true representations of the stellar absorption redshifts .
our selection of velocity - offset emission lines depends on the redshift of the galaxy s stellar absorption features , so we measure these redshifts using the high equivalent width absorption lines ca h+k , g - band , and . for each of the 8136 spectra in our sample , we constructed a rest - frame , model spectrum of the stellar continuum with the gandalf algorithm @xcite , and from the template we isolated a region of flux , covering 100 , around each of the three stellar absorption features .
we fit a redshift to each of the three regions , and we took the host galaxy redshift to be the mean of these three redshifts .
we measured the fluxes of nine emission lines (;
@xmath26@xmath28 ; @xmath26@xmath29 ; ; @xmath26@xmath30 ; @xmath26@xmath31 ) covered in the wavelength range of our sample . after subtracting a continuum model from each spectrum
, we fit gaussians to each emission feature .
we fixed the values of / @xmath26and / @xmath26according to the line flux ratios set by atomic physics , and we tied the wavelength centroids of @xmath26and @xmath26together and the wavelength centroids of @xmath26and @xmath26together .
we also tied the widths of the , , , and @xmath26profiles .
we fit both single and double gaussian profiles to each emission line , since double gaussian fits are more suitable for double - peaked profiles or systems with wide emission - line wings . for the double gaussian
fits , we applied all of the criteria described above for the primary ( narrower ) gaussian , while for the secondary ( broader ) gaussian we tied the wavelength centroids to each other , the gaussian widths to each other , and the ratios of the primary gaussian flux to the secondary gaussian flux to each other .
we defined a double gaussian fit as more appropriate for a spectrum when the following relation @xcite applies for all six emission lines : @xmath32 where @xmath33 and @xmath34 are the chi - squared values for the single gaussian fits and the double gaussian fits , respectively , to the emission lines .
we measured the emission line fluxes as the areas under the best - fit gaussians , with uncertainties propagated from the errors on the fitting parameters .
ccccccccc & 0.2675 & @xmath35 & @xmath36 & @xmath37 & @xmath38 & @xmath39 & @xmath40 & @xmath41 + j142452.14 + 343209.4 & 0.2613 & @xmath42 & @xmath43 & @xmath44 & @xmath45 & @xmath46 & @xmath47 & @xmath48 + j142505.76 + 324733.2 & 0.3694 & @xmath49 & @xmath50 & @xmath51 & @xmath52 & @xmath53 & @xmath54 & @xmath55 + j142506.00 + 350927.5 & 0.1921 & @xmath56 & @xmath57 & @xmath58 & @xmath59 & @xmath60 & @xmath61 & @xmath62 [ tbl : agn ] since we will use the , , , and @xmath26emission lines to select agns ( see section [ agn ] ) , it is essential to a clean selection that these emission lines are significantly detected .
to remove noisy spectra , we required that each of the four emission lines is detected with at least 2@xmath63 significance .
we also required an @xmath26equivalent width @xmath64 5 , where we measured equivalent width as the ratio of the @xmath26flux to the median flux in the continuum near .
ages observed some galaxies more than once , creating duplicate spectra in the catalog . in cases
where both duplicate spectra passed the quality cuts described above , we retained the spectrum with the smaller chi - squared value for the fit to the emission lines ( section [ emission ] ) for the rest of our analysis . after the quality cuts and removal of duplicates , our sample consists of high quality spectra of 4481 galaxies , or 55@xmath65 of the original catalog at @xmath0 . using our measured emission line fluxes ( section [ emission ] ) ,
we selected agns with the standard baldwin phillips
terlevich ( bpt ) diagram of line ratios ( @xcite ; figure [ fig : bpt ] ) .
our selection of galaxies with line flux ratios above the theoretical agn
starburst boundary @xcite yielded 182 agn candidates .
we inspected these 182 candidates for quality by eye and found that one has poorly subtracted night sky lines and eight have artificially rising fluxes towards the red end of the spectra , which call into question the accuracy of the fits to their line fluxes and subsequent selection as agns . after removing these nine objects ,
our sample consists of 173 agns ( table [ tbl : agn ] ) .
all of the agns are type 2 agns , since the @xmath26and @xmath26emission lines are well reproduced by a model that ties their widths to the @xmath26and @xmath26line widths without an additional broad component ( section [ emission ] ) .
the / @xmath26and / @xmath26line flux ratios can also be used to distinguish seyferts , low - ionization narrow emission - line regions ( liners ) where the power sources are still unclear ( e.g. , @xcite ) , and ambiguous cases that do not have uniform classifications across the line diagnostics .
the line flux ratios of the 173 agns show that 82 are seyferts , 22 are liners , and 69 are ambiguous @xcite . of the 69 ambiguous cases ,
26 have line flux ratios that are , within their errors , consistent with seyfert classifications .
therefore , a total of 108 agns ( @xmath66 of the population ) are either classified as seyferts or ambiguous systems that are consistent with seyfert classifications .
we proceed with our analysis of the 173 agns , noting that the majority of the ages agn are likely seyferts and that the other studies to which we will compare ( section [ compare ] ) used agn classifications without separations for liners and ambiguous cases . for this smaller sample of 173 type 2 agns we redid the fits to the continuum - subtracted emission lines and allowed the velocities of the emission lines to float relative to each other , since in principle different lines may display different redshifts ( e.g. , @xcite ) .
as in section [ emission ] , we applied both single and double gaussian fits to the lines . in the case of a single gaussian or a primary ( narrower ) component of a double gaussian , we set the wavelength centroids of the , , and @xmath26features to be free parameters , while we tied the @xmath26wavelength centroids to that of .
we also tied the wavelength centroids of all of the secondary ( broader ) gaussians to one another .
the median full widths at half maximum of the best fit narrow and broad components are 6 and 14 , respectively .
we used the chi - squared criterion ( equation [ chi - squared ] ) to determine whether a single or double gaussian is the more suitable fit for each spectrum . in some instances we found that a double gaussian fit is appropriate for the @xmath26lines while a single gaussian fit is appropriate for the , , and @xmath26lines , as has been seen in other agn samples ( e.g. , @xcite ) .
( top ) , @xmath26(middle ) , and @xmath26(bottom ) relative to the host galaxy stellar absorption features for our sample of 173 type 2 agns in ages .
the colored segments denote the velocity offsets of the five offset agn candidates and the two dual agn candidates , as selected in section 3 ( see figure [ fig : offset ] for color identifications ) .
the velocity offsets of both the redward and blueward peaks of the double - peaked agns are shown.,width=340 ] to determine the robustness of the fits to the emission lines , we added noise to each agn spectrum and redid the fits described in section [ rigorous ] . using monte carlo realizations , we drew the noise 100 times from a gaussian distribution with the variance of the spectrum s flux and redid the emission line fits each time .
the noise is dominated by variance in the sky subtraction , while systematics in flat - fielding and flux calibration are unknown and not accounted for here . for each realization
we measured the flux of each emission line as the area under the gaussian fit , and we used the mean and standard deviation of these values as the line s flux and uncertainty .
we also used the host galaxy redshifts to convert these line fluxes to luminosities .
the gaussian fit to each line also yields a wavelength centroid , and in the case of a double gaussian fit we used the central wavelength of the primary ( narrower ) gaussian as the line s representative wavelength .
we measured the wavelength of each emission line as the mean and standard deviation of the wavelength centroids determined from the 100 realizations of adding noise to each spectrum .
we then combined the wavelengths of the emission lines and the wavelengths of the stellar absorption features ( measured in section [ stellar ] ) to determine the line - of - sight velocity offset , @xmath67 , of each agn emission line relative to the host galaxy stellar absorption features .
when we compare the velocity offsets of the emission lines in each galaxy , we find that the median velocity difference of @xmath26 ( ) relative to @xmath26is @xmath68 km s@xmath10 ( @xmath69 km s@xmath10 ) . figure [ fig : hist ] shows histograms of the velocity offsets @xmath67 measured for , , and . to obtain an overall averaged velocity offset for each system , we took the mean of the three velocity offset measurements weighted by their inverse variances .
we examined the 173 agn spectra by eye and selected 20 spectra as candidates for double - peaked emission lines . to determine which spectra are truly double - peaked , and hence dual agn candidates , we fit double gaussian profiles to the emission lines in each of the 20 spectra .
we selected the double - peaked agns as those that have / @xmath26and / @xmath26flux ratios that indicate that both the redward and the blueward emission components are produced by agns @xcite , which yielded two double - peaked agns ( figure [ fig : offset ] ; table [ tbl : properties ] ) . using the inverse - variance - weighted mean velocity offsets of the red peaks and blue peaks
, we find the line - of - sight velocity separations of the two double - peaked agns are @xmath70 km s@xmath10 for ndwfs j143208.27 + 353255.5 and @xmath71 km s@xmath10 for ndwfs j143359.71 + 351020.5 . at the redshifts of these galaxies , the minimum velocity separations resolvable with the ages spectral resolution are 253 km s@xmath10 and 205 km s@xmath10 , respectively .
cccccccc & 0.0834 & @xmath72 & @xmath73 & @xmath74 & @xmath75 & @xmath76 & seyfert + & & @xmath77 & @xmath78 & @xmath79 & @xmath80 & @xmath81 & seyfert + ndwfs j143359.71 + 351020.5 & 0.3372 & @xmath82 & @xmath83 & @xmath84 & @xmath85 & @xmath86 & liner + & & @xmath87 & @xmath88 & @xmath89 & @xmath90 & @xmath91 & seyfert + & 0.2297 & @xmath92 & @xmath93 & 0.75 @xmath94 0.03 & @xmath95 & @xmath96 & liner + ndwfs j143053.69 + 345836.4 & 0.0839 & @xmath97 & 3.39 @xmath94 0.10 & 0.86 @xmath94 0.02 & @xmath98 & @xmath99 & seyfert + ndwfs j143316.48 + 353259.3 & 0.1994 & @xmath100 & 3.10 @xmath94 0.36 & 0.41 @xmath94 0.04 & @xmath101 & @xmath102 & seyfert + ndwfs j143317.07 + 344912.0 & 0.3637 & @xmath103 & 4.23 @xmath94 1.41 & 0.41 @xmath94 0.03 & @xmath104 & @xmath105 & ambiguous@xmath106 + ndwfs j143710.03 + 343530.1 & 0.1266 & @xmath107 & 2.41 @xmath94 0.14 & 0.64 @xmath94 0.03 & @xmath108 & @xmath109 & ambiguous [ tbl : properties ] . for clarity ,
the spectra are normalized in flux and offset vertically .
each spectrum is shifted to the rest wavelength of its host galaxy , based on the host galaxy redshift given by the stellar absorption features .
the dotted vertical lines show the restframe wavelengths of , , , and .
the blueward and redward components of the double gaussian fits to the double - peaked agns are shown as the blue and red curves , respectively , and the emission line velocity shifts are also visible in the offset agn candidates . , width=340 ] next , we searched the ages spectra for offset agn candidates .
velocity offsets in agn emission lines can be produced not only by offset agns , but also by other kinematic effects such as agn outflows .
an agn outflow that decelerates with distance from the central agn will produce a stratified velocity structure , imparting higher velocities to lines with higher ionization potentials , such as , and lower velocities to lines with lower ionization potentials , such as @xmath26and @xmath26(e.g . , @xcite ) .
in contrast , all of the emission lines should exhibit the same velocities if they are produced by the bulk motion of an offset agn moving within the host galaxy .
we searched the 173 agn spectra for signatures of offset agns that have bulk velocities within their host galaxies .
first , to avoid emission line systems with nonzero velocity differences that are caused by measurement errors , we selected the 14 agns with velocity offsets that are different from zero by @xmath110 for each of the , , and @xmath26emission lines ( we note that there are an additional 32 agns that have @xmath111 velocity offsets in @xmath26but not @xmath26or ; these may be agn outflows ) .
then , we selected the five agns that have , and , and @xmath26velocity offsets that are all consistent to within @xmath112 .
these five agns ( figure [ fig : offset ] ; table [ tbl : properties ] ) are our offset agn candidates .
the inverse - variance - weighted mean velocity offsets of the five offset agn candidates span @xmath113 km s@xmath10 @xmath114 km s@xmath10 ( table [ tbl : properties ] ) .
although the selection criteria are similar , we note that these velocity - offset narrow agn emission lines are a different class from the velocity - offset _ broad _ agn emission lines used in searches for subparsec - scale smbh binaries and gravitationally recoiling smbhs ( e.g. , @xcite ) .
binary and recoiling smbh searches typically focus on the broad - line region carried with the smbh at velocities @xmath115 km s@xmath10 , while our search for dual smbhs targets narrow - line velocity offsets , which are up to @xmath116few hundred km s@xmath10 for kpc - scale separation dual smbhs .
luminosity ( right ) distributions of the seven offset and dual agn candidates ( red ) as compared to the parent population of ages agns ( black ) .
both populations are consistent with being drawn from the same distribution of redshifts and the same distribution of @xmath26luminosities .
, width=340 ] to test whether the velocity offset agns have other unique observable characteristics , we compare the subset of seven offset and dual agn candidates to the parent population of 173 type 2 agns in ages .
the redshift distributions of the two samples , shown in figure [ fig : compare ] , are consistent . according to a kolmogorov - smirnov test
, there is a 66@xmath65 probability that the redshifts of the seven dual smbh candidates were drawn from the same distribution as the redshifts of the ages agn parent sample .
we also examine the @xmath26luminosities ( measured in section [ agnmeasure ] ) and find a probability of 66@xmath65 that the @xmath26luminosities of the dual smbh candidates and agn parent population were drawn from the same distribution .
finally , we note that the spectral resolutions are too low to measure resolved gas velocity dispersions for the entire sample . in general , we find that the dual smbh candidate subpopulation has redshifts and @xmath26luminosities that are similar to those of the parent agn population
. however , it is striking that the same offset agn candidate , ndwfs j143317.07 + 344912.0 , has both the second highest redshift ( @xmath117 ) and the largest @xmath26luminosity ( @xmath118 } } = 5.2 \times 10^{9 } \ , l_\odot$ ] ) of the entire sample of 173 type 2 agns in ages . we discuss additional interesting features of this system in the following section .
color and absolute @xmath119-band magnitude m@xmath120 , based on sdss photometry of the host galaxies of all 173 ages agns in our sample .
the seven offset and dual agn candidates ( red triangles ) have host galaxies with similar colors and absolute magnitudes to those of the overall agn population ( black circles ) .
, width=340 ] we also use sdss photometry to examine the host galaxies of the offset and dual agn candidates , and we find that the candidates host galaxies colors and absolute magnitudes are similar to those of the parent population of agn host galaxies ( figure [ fig : colormag ] ) . for @xmath121 colors ( m@xmath120 absolute magnitudes ) , there is a 64@xmath65 ( 72@xmath65 ) probability that the candidates were drawn from the same distribution as all agn host galaxies .
one interesting exception is that the reddest agn host galaxy ( @xmath122 ) is the host galaxy of one of the double - peaked agns , ndwfs j143359.71 + 351020.5 .
if this galaxy is shown to indeed host dual agns , it will be interesting to examine whether the galaxy is dust - reddened or whether there is a lack of ongoing star formation to reconcile with the fueling of two agns .
further , we visually inspect the sdss _ gri _ color - composite images of the host galaxies for evidence of galaxy mergers that may have caused the agn velocity offsets we measure in the seven dual smbh candidates .
the images of the two double - peaked agns are shown in figure [ fig : dual_image ] , and the images of the five offset agn candidates are shown in figure [ fig : offset_image ] . whereas only @xmath7 ( 16/173 ) of the ages agn host galaxies have a companion within @xmath123 , this figure increases to @xmath8 ( 2/7 ) for the dual smbh candidates .
the two offset agn candidates with nearby companions are ndwfs j143316.48 + 353259.3 and ndwfs j143317.07 + 344912.0 , which is also the galaxy with the second highest redshift and the largest @xmath26luminosity in the agn sample ( section [ parent ] ) . notably ,
neither companion is so close that it overlaps with the @xmath22 diameter hectospec fiber used to obtain the ages spectrum of the primary galaxy .
therefore , the stellar continuum and the agn emission features we measure in these ages spectra arise solely from the primary galaxy and not from , e.g. , a combination of continuum from the primary and agn emission from the companion .
we discuss each companion in detail in sections [ companion1 ] and [ companion2 ] .
we find evidence that dual smbh candidate host galaxies have a @xmath6 times higher probability of having companions , as compared to the general population of agn host galaxies .
this greater probability suggests that mergers may be at the root of the agn velocity offsets we measure here .
we note that in the cases of the two offset agn candidates that have companions , both the primary and companion galaxies have faint , spatially - extended features that resemble tidal tails .
this implies that both systems may be galaxy mergers in progress that have already had their first pericenter passages .
the initial collision or collisions could have produced the velocity offsets we measure in the ages spectrum by disrupting the velocity of the stars relative to the central agn ( e.g. , @xcite ) , by producing tidal features that skew the observed velocity of the stars due to projection effects , or by jostling the agn to a different velocity than the stars .
, with north up and east to the left .
the scale bars in the lower left hand corners illustrate the physical distance scales at the redshifts of the host galaxies .
, width=340 ] the southeast companion to ndwfs j143316.48 + 353259.3 is separated by 15 kpc and is 1.5 times more luminous than the primary .
the companion has an ages spectrum , and we measure this companion s redshift and its emission line fluxes and velocity offsets as in section 2 .
we note that the companion , the galaxy ndwfs j143316.70 + 353256.4 , did not survive our initial quality cut ( section [ quality ] ) .
although the emission lines are detected at @xmath124 significance , the @xmath26equivalent width of 2.3 is below our quality cutoff of 5 .
we find that the companion displays an almost identical redshift to the primary ( table [ tbl : companions ] ) , placing it at a velocity @xmath125 km s@xmath10 ( blueward ) of the primary galaxy .
based on its flux ratios of @xmath126 $ \lambda$5007}}/{\mbox{h$\beta$ } $ \,$}=1.55 \pm 0.51 $ ] and @xmath127 $ \lambda$6584}}/{\mbox{h$\alpha$ } $ \,$}=0.73 \pm 0.15 $ ] , the companion just makes the classification as a pure agn @xcite , but the errors on the line flux measurements include the possibility of an agn star formation composite system ( @xcite ; figure [ fig : companion ] ) .
the emission lines have line - of - sight velocity offsets ( @xmath128 ) of @xmath129 km s@xmath10 , @xmath130 $ \lambda$5007}}}=-22.4 \pm 29.4 $ ] km s@xmath10 , and @xmath131 km s@xmath10 .
the inverse - variance - weighted mean of the velocity offsets is @xmath132 km s@xmath10 .
the velocity offsets of the three emission lines are all consistent to within @xmath112 , suggestive of the bulk motion of an offset agn ( see section [ offset ] ) , but each velocity offset measurement is also @xmath133 from zero .
either the agn has a small line - of - sight velocity offset relative to systemic or it has no offset ; additional observations with improved spectral resolution could distinguish between these scenarios .
the similar emission - line velocity offsets in the two galaxies ( @xmath134 km s@xmath10 for the primary and @xmath135 km s@xmath10 for the companion ) and the nearly identical redshifts derived from the two sets of stellar absorption features ( the difference is @xmath136 km s@xmath10 ) suggests that the two galaxies are dynamically linked .
when combined with the tidal features visible in the sdss image , this is evidence that the system is one or more pericenter passages into the merger process . , but for the five offset agn candidates in ages . in the systems that have multiple interacting galaxies , the primary ages sources are marked with crosshairs.,width=340 ] located 16 kpc from ndwfs j143317.07 + 344912.0 is a northern companion that is 1.7 times less luminous than the primary .
although there are not ages or sdss spectra for the companion , the photometric redshifts from sdss suggest that the companion is at roughly the same redshift as the primary ( table [ tbl : companions ] ; photometric redshifts for ndwfs j143316.48 + 353259.3 are also shown for comparison , since its companion is spectroscopically confirmed ) .
however , a spectroscopic redshift is required to determine whether the companion is indeed associated with ndwfs j143317.07 + 344912.0 .
cccccc & 0.199419 @xmath94 0.0003 & @xmath137 & @xmath138 & 15 & 1.3 +
ndwfs j143316.70 + 353256.4@xmath139 & 0.199418 @xmath94 0.0003 & @xmath140 & @xmath141 & & 1.9 + & 0.3637 @xmath94 0.0003 & @xmath142 & @xmath143 & 16 & 4.7 + northern companion & n / a & @xmath144 & @xmath145 & & 2.8 [ tbl : companions ] we find two double - peaked agns out of the 173 type 2 agns at @xmath0 in ages , which corresponds to a rate of @xmath146 . for comparison ,
the fraction of double - peaked agns is @xmath147 ( 2/91 ) in the deep2 sample of type 2 agns in red galaxies at @xmath148 @xcite , @xmath149 ( 87/6780 ) in sdss type 2 agns at @xmath150 @xcite , and @xmath151 ( 167/14756 ) in sdss type 2 agns and type 2 quasars at @xmath152 @xcite . since all of the double - peaked agns in ages , deep2 , and sdss have velocity separations @xmath153 km s@xmath10 , the differences in the surveys spectral resolutions ( @xmath154 for ages , @xmath155 for deep2 , and @xmath156 for sdss ) do not bias the comparison of the fractions of double - peaked agns found in each .
the three surveys do not probe the same spatial scales ( sdss uses @xmath157 fibers , which corresponds to 5.5 kpc at typical redshift @xmath158 ; ages uses @xmath22 fibers , which corresponds to 5.0 kpc at typical redshift @xmath159 ; and deep2 uses slits with 1@xmath160 widths and 7@xmath160 average lengths , which correspond to 7 kpc and 50 kpc , respectively , at typical redshift @xmath161 of the double - peaked and offset sample ; @xcite ) , allowing the possibility that the double peaks in each survey are produced by effects on different spatial scales .
although deep2 can probe scales of 50 kpc , the sources of the double - peaked emission exist on @xmath162 kpc scales .
the two double - peaked agns in deep2 correspond to double emission components with spatial separations of 1.2 kpc and 2.3 kpc @xcite , and long - slit observations of 81 double - peaked agns in sdss showed that the double emission components have a typical separation of 1.1 kpc ( with a range of 0.2 kpc to 5.5 kpc , where the maximum is 15 kpc for a @xmath157 sdss fiber ; @xcite ) . as a result , the different spatial scales explored by sdss , ages , and deep2 should not significantly bias the relative numbers of offset and dual agn candidates found in each . finally , the deep2 search was limited to red galaxies . if we match the ages agn sample to the color ( @xmath163 ) and absolute magnitude ( @xmath164 ) range of the deep2 sample , we find zero double - peaked system out of 70 agns .
we also find five offset agn candidates , which is @xmath165 of our sample of 173 type 2 agns at @xmath0 . for comparison ,
the fraction is @xmath166 ( 30/91 ) for offset agn candidates in the deep2 sample of type 2 agns in red galaxies at @xmath148 @xcite .
a major source of this discrepancy is that the deep2 spectra have five times the spectral resolution ( @xmath155 ) of the ages spectra ; the deep2 offset agn candidates have velocity offsets down to 37 km s@xmath10 with @xmath111 significance , whereas the ages offset agn candidates have velocity offsets down to 98 km s@xmath10 with @xmath111 significance .
a more accurate comparison is the @xmath167 ( 14/91 ) fraction of deep2 type 2 agns that have velocity offsets @xmath168 km s@xmath10 to the @xmath169 ( 2/70 ) fraction of offset agn candidates in an ages agn sample that is matched in color and absolute magnitude ( as described in the previous paragraph ) to the deep2 sample .
when the ages and deep2 samples are matched in color , absolute magnitude , and minimum velocity offset , there is clear evidence for redshift evolution in the fractions of offset and dual agn candidates . for the subsamples matched in this way
, the fraction of offset ( dual ) agn candidates increases with redshift , from @xmath170 ( @xmath171 ) in ages at @xmath0 and mean redshift @xmath14 , to @xmath172 ( @xmath173 ) in deep2 at @xmath174 and mean redshift @xmath16 . when the offset and dual agn candidates are combined , the overall fractions increase by a factor of @xmath3 from @xmath4 ( 2/70 ) in ages to @xmath5 ( 16/91 ) in deep2 .
this trend is similar to the increase in galaxy merger fraction ( e.g. , @xcite ) with increasing redshift over this redshift range , as expected if these candidates are indeed offset and dual agns .
if the five ages candidates are offset agns , then we expect there to be equal numbers of redshifted and blueshifted agns , as was the case for the offset agn candidates in deep2 @xcite . in ages ,
@xmath175 ( 1/5 ) of the candidates exhibit redshifted velocity offsets and @xmath176 ( 4/5 ) of the candidates exhibit blueshifted velocity offsets .
given the small sample size , this could be roughly consistent with an even distribution , or it could be an indication of agn outflows biasing the sample .
agn outflows are known to produce an overabundance of observed blueshifted emission lines , since redshifted lines are often obscured by the agn torus ( e.g. , @xcite ) .
we have searched 8136 ages galaxy spectra at @xmath0 for spectroscopic signatures of dual smbhs .
dual smbhs that are accreting gas as dual agns can produce double - peaked agn emission lines , while an offset agn , which is a dual - smbh system where only one smbh is active , can produce agn emission lines with bulk line - of - sight velocity offsets relative to the host galaxy stellar absorption features . out of 173 type 2 agn spectra at @xmath0 , we find two double - peaked agns and five offset agn candidates .
the two double - peaked agns have line - of - sight velocity separations between their double peaks of 246 km s@xmath10 and 358 km s@xmath10 , while the five offset agn candidates have velocity offsets of @xmath113 km s@xmath10 @xmath114 km s@xmath10 .
we find that @xmath146 of the ages agns are double - peaked agns and @xmath165 are offset agn candidates . when the ages and deep2 samples are matched in color , absolute magnitude , and minimum velocity offset , the dual supermassive black hole candidate fraction increases by a factor of @xmath3 ( from 2/70 to 16/91 , or @xmath4 to @xmath5 ) from the ages sample at a mean redshift of @xmath14 to the deep2 sample at @xmath16 .
this trend could be understood if velocity - offset narrow agn emission lines are associated with galaxy mergers , since the galaxy merger fraction also increases with redshift over this range . in an additional sign of a link between velocity offsets and galaxy mergers ,
we find tentative evidence that the offset and dual agn candidates are @xmath6 times more likely ( 2/7 , or @xmath8 ) to be hosted by galaxies with a companion within @xmath123 than the parent ages agn population ( 16/173 , or @xmath7 ) .
since our sample consists of only seven offset and dual agn candidates , a larger sample is needed to confirm this result .
further , we find that two of the offset agn candidates have companions , and one of the companions may host an offset agn itself that has a similar velocity offset to the primary galaxy s agn
. both sets of primary and companion galaxies also have morphological signs of disturbance , suggesting that they have already interacted . in the dual smbh scenario ,
such an interaction could perturb the central agn to a different velocity than systemic . while the seven objects we find are compelling candidates for offset and dual agns , their line profiles could also be produced by kinematics in the narrow - line region such as outflows and rotating disks ( e.g. , @xcite ) .
follow - up studies of the sdss double - peaked agns have showcased the utility of long - slit spectroscopy @xcite , integral - field unit spectroscopy @xcite , near - infrared imaging @xcite , radio observations @xcite , and x - ray observations @xcite in distinguishing narrow - line region kinematics from bona fide dual agns . while radio and x - ray observations are particularly promising avenues for confirmations of dual agns , the _ chandra _ observations to date have underscored the difficulties in identifying dual agns when the agns are closely separated and faint or have high x - ray obscurations @xcite .
building on the strategies honed for sdss double - peaked agns , careful follow - up observations of the ages offset and dual agn candidates presented here would pinpoint which systems are dual smbhs , which would help clarify our understanding of spectroscopic signatures of dual smbhs and advance these systems as new probes of galaxy evolution .
, r. s. , stern , d. , madsen , k. , harrison , f. , assef , r. j. , comerford , j. m. , cushing , m. c. , fassnacht , c. d. , gonzalez , a. h. , griffith , r. , hickox , r. , kirkpatrick , j. d. , & lagattuta , d. j. 2012 , , 744 , 7 , b. f. , newman , j. a. , lotz , j. , yan , r. , barmby , p. , coil , a. l. , conselice , c. j. , ivison , r. j. , lin , l. , koo , d. c. , nandra , k. , salim , s. , small , t. , weiner , b. j. , cooper , m. c. , davis , m. , faber , s. m. , & guhathakurta , p. 2007
, , 660 , l23 , l. , strauss , m. a. , tremonti , c. a. , schlegel , d. j. , heckman , t. m. , kauffmann , g. , blanton , m. r. , fan , x. , gunn , j. e. , hall , p. b. , ivezi , . , knapp , g. r. , krolik , j. h. , lupton , r. h. , richards , g. t. , schneider , d. p. , strateva , i. v. , zakamska , n. l. , brinkmann , j. , brunner , r. j. , & szokoly , g. p. 2005
, , 129 , 1783 , b. t. , & dey , a. 1999 , in asp conf .
191 , photometric redshifts and the detection of high redshift galaxies , ed r. weymann , l. storrie - lombardi , m. sawicki , and r. brunner ( san francisco , ca : asp ) , 111 , g. , heckman , t. m. , tremonti , c. , brinchmann , j. , charlot , s. , white , s. d. m. , ridgway , s. e. , brinkmann , j. , fukugita , m. , hall , p. b. , ivezi , . , richards , g. t. , & schneider , d. p. 2003
, , 346 , 1055 , c. s. , eisenstein , d. j. , cool , r. j. , caldwell , n. , assef , r. j. , jannuzi , b. t. , jones , c. , murray , s. s. , forman , w. r. , dey , a. , brown , m. j. i. , eisenhardt , p. , gonzalez , a. h. , green , p. , & stern , d. 2012 , , 200 , 8 , l. , patton , d. r. , koo , d. c. , casteels , k. , conselice , c. j. , faber , s. m. , lotz , j. , willmer , c. n. a. , hsieh , b. c. , chiueh , t. , newman , j. a. , novak , g. s. , weiner , b. j. , & cooper , m. c. 2008 , , 681 , 232 , m. , falcn - barroso , j. , davies , r. l. , bacon , r. , bureau , m. , cappellari , m. , de zeeuw , p. t. , emsellem , e. , fathi , k. , krajnovi , d. , kuntschner , h. , mcdermid , r. m. , & peletier , r. f. 2006 , , 366 , 1151 | dual supermassive black holes ( smbhs ) with kiloparsec scale separations in merger - remnant galaxies are informative tracers of galaxy evolution , but the avenue for identifying them in large numbers for such studies is not yet clear .
one promising approach is to target spectroscopic signatures of systems where both smbhs are fueled as dual active galactic nuclei ( agns ) , or where one smbh is fueled as an offset agn .
dual agns may produce double - peaked narrow agn emission lines , while offset agns may produce single - peaked narrow agn emission lines with line - of - sight velocity offsets relative to the host galaxy .
we search for such dual and offset systems among 173 type 2 agns at @xmath0 in the agn and galaxy evolution survey ( ages ) , and we find two double - peaked agns and five offset agn candidates .
when we compare these results to a similar search of the deep2 galaxy redshift survey and match the two samples in color , absolute magnitude , and minimum velocity offset , we find that the fraction of agns that are dual smbh candidates increases from @xmath1 to @xmath2 by a factor of @xmath3 ( from 2/70 to 16/91 , or @xmath4 to @xmath5 ) . this may be associated with the rise in the galaxy merger fraction over the same cosmic time . as further evidence for a link with galaxy mergers ,
the ages offset and dual agn candidates are tentatively @xmath6 times more likely than the overall agn population to reside in a host galaxy that has a companion galaxy ( from 16/173 to 2/7 , or @xmath7 to @xmath8 ) .
follow - up observations of the seven offset and dual agn candidates in ages will definitively distinguish velocity offsets produced by dual smbhs from those produced by narrow - line region kinematics , and will help sharpen our observational approach to detecting dual smbhs . |
we continue our development on algorithms to count the points of various representation varieties of a quiver with relations . in @xcite
, we applied serval counting characters to the harder - narasimhan identity in the hall algebra of a quiver and obtained several interesting formulas .
all characters that we considered are originated from reineke s counting character @xmath3 from the hall algebra to certain quantum power series ring .
unfortunately @xmath3 fails to be an algebra morphism for non - hereditary algebras , though harder - narasimhan identity exists quite generally .
however , applying the same map @xmath3 to the hn - identity can still generate effective counting formulas .
we will follow the similar line as the first paper .
the only change is that we replace algebraic manipulations in the hall algebras by corresponding geometric constructions .
we first state the main results of this notes .
let @xmath4 to be the finite field @xmath0 with @xmath5 elements and @xmath6 be any basic algebra presented by @xmath7 .
fix a slope function @xmath8 , and we denote by @xmath9 the variety of @xmath10-dimensional @xmath8-semistable representations of @xmath6 , and by @xmath11 its git quotient .
@xmath12 where the sum runs over all decomposition @xmath13 of @xmath10 into non - zero dimension vectors such that @xmath14 for @xmath15 .
we will define the key varieties @xmath16 in section 1 .
in particular , if all @xmath17 varieties can be effectively counted , then so are @xmath9 .
the map @xmath3 have so - called @xmath18 and @xmath19 analogs .
they are defined in @xcite as @xmath20 and @xmath21 . here , @xmath18 and @xmath19
are related to the comultiplication and the antipode @xcite in the hall algebra . in our geometric setting ,
lemma 0.1 and @xmath17 varieties have @xmath18 and @xmath19 analogs as well .
recall that a variety @xmath22 is called _ polynomial - count _ ( or has a counting polynomial ) if there exists a ( necessarily unique ) polynomial @xmath23 $ ] such that for every finite extension @xmath24 , we have @xmath25 .
we are especially interested in when all @xmath17 varieties are polynomial - count .
if this is the case , it is clear that each @xmath11 is polynomial - count when it is a geometric quotient . in this notes
, we will mainly focus on a class of algebras called one - point extensions from a quiver .
let @xmath1 be any finite quiver and @xmath26 a representation of @xmath1 .
the one - point extension of @xmath1 by @xmath26 is the triangular algebra @xmath27:={\left(\begin{smallmatrix}kq\ 0\\\,e\:\ k\end{smallmatrix}\right)}$ ] .
we also interested in the tensor product algebra @xmath28 , where @xmath29 is the dynkin quiver of type @xmath29 . for @xmath30 $ ] or @xmath31
, we have explicit counting formulas for @xmath17 varieties and their @xmath18 and @xmath19 analogs .
[ t : intro ] 1 . for @xmath30 $ ] ,
all git quotients @xmath11 can be explicitly counted in terms of quiver grassmannians of @xmath26 .
if @xmath26 is add - polynomial - count , then all @xmath11 are polynomial - count .
2 . for @xmath32 ,
all @xmath11 have counting polynomials , which can be explicitly computed .
3 . if @xmath26 is add - polynomial - count , @xmath33)$ ] is polynomial - count for certain choice of @xmath10 and @xmath8 .
this notes are organized as follows . in section [ s :
prelim ] , we provide necessary background on the representation theory of quivers with relations and points counting . in section [ s : hn ] , we introduce the @xmath17 variety and the notion of f - polynomial - count . after recalling the harder - narasimhan identity in the hall algebra
, we conclude our key lemma ( lemma [ l : tao ] ) . in section [ s :
ext ] , we first review the trivial extension of algebras in general , then specialize to the case of one - point extensions from a quiver .
we describe the relations of these algebras from the projective presentation of @xmath26 . in section [ s :
frep ] , we show in lemma [ l : frep ] that their @xmath17 varieties can be counted in terms of the usual representation varieties . in section [ s : rep ] , we show in lemma [ l : rep ] that these usual representation varieties can be counted in terms of the grassmannians of @xmath34 . motivated by this , we introduce add - polynomial - count property for a representation .
we conclude by our fist main result theorem [ t : ext ] ( theorem [ t : intro].(1 ) ) .
many examples will follow in section [ s : example ] . in section [ s :
hs ] , we apply our algorithm to count homological strata on the geometric quotients .
the method is outlined in theorem [ t : hs ] . in section
[ s : a2 ] , we work with the algebra @xmath35 . our second main result theorem [ t : a2q ] ( theorem [ t : intro].(2 ) ) gives an analogous counting formula , which is independent of grassmannians of representations .
in section [ s : delta ] , we consider the @xmath18-analog of counting .
we introduce the @xmath18-analog of the @xmath17 varieties .
lemma [ l : frep2 ] is the @xmath18-analog of lemma [ l : frep ] .
we conclude by our third main results
theorem [ t : a2ext ] ( theorem [ t : intro].(3 ) ) . finally in section
[ s : s ] , we consider the @xmath19-analog of counting .
our final main results is theorem [ t : final ] , which removes the assumption of being a geometric quotient in our previous results .
most of our constructions can be easily generalized to the motivic setting . since the main application of this theory will be in the quantum algebra
, we will not pursue that generality .
the geometry of these moduli spaces will be studied in another series of notes @xcite .
let @xmath1 be a finite quiver with the set of vertices @xmath36 and the set of arrows @xmath37 .
if @xmath38 is an arrow , then @xmath39 and @xmath40 denote its tail and its head respectively .
fix a _ dimension vector _ @xmath10 , the space of all @xmath10-dimensional representations over a field @xmath4 is @xmath41 the group @xmath42 acts on @xmath43 by the natural base change .
two representations @xmath44 are isomorphic if they lie in the same @xmath45-orbit .
let @xmath46 be the _ path algebra _ of @xmath1 over @xmath4 , then @xmath47 is naturally a ( right ) @xmath46-module .
fix a set @xmath48 of homogeneous elements in @xmath46 with respect to the bigrading : @xmath49 here , we abuse the vertex @xmath50 for the trivial path @xmath51 .
if @xmath52 for all @xmath53 , then we say @xmath54 is a _ representation of @xmath1 with relations @xmath48_. the _ path algebra of @xmath1 with relations @xmath48 _ is the quotient algebra @xmath55 . a representation of @xmath1 with relations
@xmath48 naturally becomes an @xmath6-module .
the assignment @xmath56 defines a polynomial map @xmath57 which is represented by an @xmath58 matrix with entries in @xmath59 $ ] .
let @xmath60 $ ] be the ideal generated by the entries of all @xmath61 for which @xmath53 .
the representation space @xmath62 is the scheme @xmath63/\tilde{r})$ ] . as a variety , @xmath62 consists of all @xmath10-dimensional representations of @xmath6 . a _ weight _ @xmath64 is an integral linear functional on @xmath65 . a _ slope function _
@xmath8 is a quotient of two weights @xmath66 with @xmath67 for any dimension vector @xmath10 .
a representation @xmath54 is called _ @xmath8-semi - stable ( resp .
@xmath8-stable ) _ if @xmath68 ( resp .
@xmath69 ) for every non - trivial subrepresentation @xmath70 .
we denote by @xmath9 the variety of @xmath10-dimensional @xmath8-semistable representations of @xmath6 . by the standard git construction @xcite
, there is a _ categorical quotient _ @xmath71 and its restriction to the stable representations @xmath72 is a _
geometric quotient_. a slope function @xmath8 is called _ coprime _ to @xmath10 if @xmath73 for any @xmath74 .
so if @xmath8 is coprime to @xmath10 , then there is no strictly semistable ( semistable but not stable ) representation of dimension @xmath10 . in this case , @xmath11 must be a geometric quotient .
note that the semi - stable objects with a fixed slope @xmath75 form an exact subcategory @xmath76 . for any dimension vector @xmath10
, we can always modify @xmath8 to get a new slope function @xmath77 such that @xmath78 and @xmath79 . if @xmath80 , then we can take @xmath81 , where @xmath82 . _
* proposition 3.3 ) _ 1 .
_ harder - narasimhan filtration : _ every representation @xmath54 has a unique filtration @xmath83 such that @xmath84 is @xmath8-semi - stable and @xmath85 .
_ jordan - holder filtration : _ every @xmath8-semi - stable representation @xmath54 has a filtration @xmath83 such that @xmath84 is @xmath8-stable with the same slope .
the set @xmath86 is uniquely determined .
+ let @xmath22 be a variety over @xmath87 and @xmath88 .
we denote by @xmath89 the @xmath90-th @xmath91-adic cohomology group with compact support of @xmath92 . the key method for counting rational points on @xmath22
is given by the grothendieck - lefschetz trace formula : @xmath93 where @xmath94 is the frobenius morphism @xmath95 .
@xmath22 is called _
@xmath91-pure _ if the eigenvalues of @xmath94 on @xmath89 have absolute value @xmath96
. it is known that if @xmath22 is smooth and proper over @xmath97 then @xmath22 is @xmath91-pure .
in fact , the condition of being proper can be weakened as follows .
* proposition a.2 ) assume that @xmath22 is smooth quasi - projective and that there is an action @xmath98 such that for every @xmath99 the limit @xmath100 exists .
assume in addition that @xmath101 is projective , then z is @xmath91-pure . _
* proposition 6.1 ) _ [ l : euler ] if @xmath22 is counted by a rational function @xmath102 , then @xmath102 must lie in @xmath103 $ ] .
its specialization at @xmath104 gives the @xmath91-adic euler characteristic of @xmath92 .
the _ poincar polynomial _
@xmath105 $ ] of @xmath22 is @xmath106 ( * ? ? ? * lemma a.1 ) [ l : polycount ] assume that @xmath22 is @xmath91-pure and polynomial - count
. then @xmath107 .
in particular , @xmath108 $ ] .
let @xmath6 be any basic algebra presented by @xmath7 for @xmath4 the finite field @xmath0 . for any decomposition of dimension vector @xmath109 , we define @xmath110 , where @xmath111 is the usual flag variety parameterizing flags of subspaces of dimension @xmath112 in @xmath113 . to simplify the notation
, we denote @xmath114 . [
d : frep ] we define the _ frep _ variety : @xmath115 let @xmath116 be the projection , the _ flag variety _ @xmath117 of @xmath54 is the fibre @xmath118 , and its subvarieties @xmath119 is @xmath120 when the flag is only 2-step , we may use the usual grassmannian notation . for example ,
@xmath121 and @xmath122 , where @xmath123 .
for any three @xmath6-modules @xmath124 and @xmath125 with dimension vector @xmath126 and @xmath127 , the _ hall number _ @xmath128 is by definition @xmath129 . for any module @xmath54 , we denote @xmath130 let @xmath131 be the space of all formal ( infinite ) linear combinations of isomorphism classes @xmath132 $ ] in @xmath6-@xmath133 .
* proposition 1.1 ) the completed _ hall algebra _ @xmath131 is the associative algebra with multiplication @xmath134[v]:=\sum_{[w]}f_{uv}^w[w],\ ] ] and unit @xmath135 $ ] .
we fix a slope function @xmath8 . for a dimension vector @xmath10 ,
let @xmath136 $ ] and @xmath137 $ ] .
our convention is that they contain zero representation @xmath138 $ ] .
we consider a simple counting map @xmath139 defined by @xmath132\mapsto a_m^{-1}$ ] . since @xmath140 , we have that @xmath141 .
we denote the function @xmath142 by @xmath143 . in general
, this function may not be rational in @xmath5 .
the existence of the harder - narasimhan filtration yields the following identity in the hall algebra @xmath131 .
* proposition 4.8 ) [ l : hnid ] @xmath144 where the sum running over all decomposition @xmath13 of @xmath10 into non - zero dimension vectors such that @xmath145 . in particular , solving recursively for @xmath146 , we get @xmath147 where the sum runs over all decomposition @xmath13 of @xmath10 into non - zero dimension vectors such that @xmath14 for @xmath15 .
the key observation is that @xmath148 so the problem boils down to counting these @xmath17 varieties . in this paper
, we will only focus on a class of algebras , for which these varieties can be effectively counted .
[ d : fpc ] we say an algebra @xmath6 is _ polynomial - count _ if each @xmath149 is polynomial - count .
it is called _ f - polynomial - count _ if each @xmath16 is polynomial - count .
we do not know a single example where @xmath6 is polynomial - count but not f - polynomial - count .
we conjecture that if each @xmath150 is polynomial - count , then @xmath6 is f - polynomial - count .
[ l : tao ] @xmath151 in particular , if @xmath6 is f - polynomial - count , then each @xmath11 is polynomial - count when it is a geometric quotient .
the assumption of being a geometric quotient in the lemma can be dropped .
given two finite - dimensional @xmath4-algebras @xmath152 and a @xmath6-@xmath153-bimodule @xmath26 , we get the _ trivial extension algebra _ @xmath154={\left(\begin{smallmatrix}b & 0\\ _ ae_b & a\end{smallmatrix}\right)}$ ] . in the meanwhile
, we can form the category @xmath155 of representations of the bimodule @xmath156 as follows : the objects are triples @xmath157 . a morphism from @xmath158 to @xmath159
is a pair @xmath160 making the following diagram commutes @xmath161^{\varphi } \ar[d]_{f_a\otimes{\operatorname{id } } } & m_b\ar[d]_{f_b}\\ n\otimes_a e \ar[r]^{\varphi } & n_b } \ ] ] ( * ? ? ?
* a.2.7 ) the two categories @xmath162)$ ] and @xmath155 are equivalent .
the equivalence is given by @xmath163 , where @xmath164 and @xmath165 .
in particular , if @xmath166)$ ] corresponds to @xmath158 , then the dimension vector of @xmath54 is given by @xmath167 .
one particular case we interested in is when @xmath26 is simply a right @xmath153-module .
think @xmath26 as a @xmath4-@xmath153-bimodule , the triangular algebra @xmath168 is called ( trivial ) _ one - point extension _ of @xmath153 by @xmath26 .
there is an obvious dual notion of one - point coextension @xmath169:={\left(\begin{smallmatrix}k & 0\\ e^ * & b\end{smallmatrix}\right)}$ ] .
[ l : variety ] 1 .
@xmath170)$ ] is the subvariety of @xmath171 defined as @xmath172 2 .
@xmath173)$ ] is the subvariety of @xmath174 defined as @xmath175 suppose that @xmath176 is presented by @xmath177 with @xmath178 and @xmath179 , where @xmath180 is the indecomposable projective representation corresponding to the vertex @xmath50
. then the algebra @xmath30 $ ] can be presented by a new quiver @xmath181 , which is obtained from @xmath1 by adjoining a new vertex
@xmath182 " and for each @xmath180 in @xmath183 a new arrow from @xmath182 " to the vertex @xmath50 .
the relations are clearly given by the matrix @xmath184 .
in reality , the presentation is always chosen to be minimal . by abuse of notation
, we also use @xmath185 $ ] to denote the new quiver @xmath181 with those new relations .
the one - point coextension @xmath186 $ ] can be similarly described using injective presentation of @xmath26 . by convention ,
the newly adjoined vertex is denoted by + " .
it is clear that @xmath26 is the first syzygy of @xmath187 , so @xmath188 .
moreover , a simple representation of @xmath185 $ ] is either @xmath187 or a simple representation of @xmath1 .
so we conclude that @xmath27 $ ] has global dimension @xmath189 .
it is easy to compute the matrices @xmath190 .
let @xmath191 , then @xmath192 so the euler matrix @xmath193 of @xmath6 is @xmath194 , where @xmath195 and @xmath196 is the euler matrix of @xmath1 . throughout this notes , @xmath197 is the multiplicative euler form of the quiver @xmath1 , that is , @xmath198 .
similarly , we define @xmath199 for @xmath200 .
to simplify our notation , we always use letter with tilde to indicate that @xmath201)$ ] can be represented by @xmath202 , where @xmath203 and @xmath204 . a dimension vector with tilde ,
say @xmath205 , consists of two components @xmath206 , or @xmath207 for coextension .
[ l : frep ] @xmath208)\to { \operatorname{fl}}_{{\tilde{\beta},\tilde{\gamma}}}$ ] is a fibre bundle with fibre @xmath209)\times { \operatorname{rep}}_{\tilde{\beta}}(q[e])\times \prod_{a\in q_1}{\operatorname{hom}}(k^{\beta(ta)},k^{\gamma(ha)})\ ] ] so @xmath210):=\frac{|{\operatorname{frep}}_{{\tilde{\beta},\tilde{\gamma}}}(q[e])|}{|{\operatorname{gl}}_{\tilde{\alpha}}| } = { \langle\beta,\gamma\rangle}^{-1}\bigl [ \begin{smallmatrix } \alpha_- \\ \gamma_-\end{smallmatrix}\bigr]|{\operatorname{gl}}_{\beta_-}|r_{\tilde{\beta}}(q[e])r_{(\alpha_-,\gamma)}(q[e]).\ ] ] we will sketch the fibre bundle construction by a picture . after fixing an elements in @xmath211 , we need to fill in the missing part for a @xmath212-dimensional representation of @xmath185 $ ] .
the missing part consists of a @xmath213-dimensional representation @xmath19 , a @xmath205-dimensional representation @xmath214 , and a bunch of linear maps from @xmath215 to @xmath216 , as indicated below .
@xmath217 we can stuff the space in the order below independently .
the linear maps from @xmath218 together with all representations @xmath19 can be identified with @xmath219)$ ] ; all representations @xmath214 can be identified with @xmath220)$ ] , and the rest of the linear maps are @xmath221 . for the last formula , we only need to notice that @xmath222}|{\operatorname{gl}}_{\alpha}|}{|{\operatorname{gl}}_{\beta}||{\operatorname{gl}}_{\gamma}|{\langle\beta,\gamma\rangle}_0}.\ ] ] the above 2-step case can be recursively generalized to the @xmath223-step case .
we only state the analog for the last formula . @xmath224)=\prod_{i=1}^s { \bigl[\begin{smallmatrix}\dot{\alpha}_{i,- } \\
\alpha_{i,- } \end{smallmatrix}\bigr]}|{\operatorname{gl}}_{\dot{\alpha}_{i-1,-}}|r_{(\dot{\alpha}_{i,-},\alpha_i)}(q[e]).\ ] ] we also state this formula for the dual case .
@xmath225)=\prod_{i=2}^s { \bigl[\begin{smallmatrix}\dot{\alpha}_{i,+ } \\
\alpha_{i,+}\end{smallmatrix}\bigr]}|{\operatorname{gl}}_{\dot{\alpha}_{i-1,+}}|r_{(\alpha_i,\dot{\alpha}_{i,+})}(q^\circ[e]).\ ] ] now the problem boils down to count those affine representation varieties @xmath226)$ ] .
for any dimension vector @xmath227 , we denote by @xmath228 the variety parameterizing all @xmath227-dimensional quotient representations of @xmath26 , and define @xmath229 [ l : rep ] @xmath230 is a fibre bundle with fibre @xmath231 so @xmath232):=\sum_{\alpha=\gamma+\beta}\frac{|{\operatorname{gr}}^{\beta}(ne)|}{{\langle\gamma,\beta\rangle}|{\operatorname{gl}}_n|}r_{\gamma}(q).\ ] ] dually , we have a formula for the one - point coextension : @xmath233)=\sum_{\alpha=\gamma+\beta}\frac{|{\operatorname{gr}}_{\gamma}(ne)|}{{\langle\gamma,\beta\rangle}|{\operatorname{gl}}_n|}r_{\beta}(q).\ ] ] the fibre bundle construction is not hard to verify .
since we only need the last formula , we give a hall algebra proof for that .
we denote by @xmath234 and @xmath235 the set of all monomorphisms and epimorphisms from @xmath54 to @xmath236 respectively .
fix a representation @xmath125 , then the following identities clearly holds in the hall algebra @xmath237}[u]\big)\big(\sum_{[v]}|{\operatorname{epi}}_q(m , v)|[v]\big ) = \sum_{[w]}|{\operatorname{hom}}_q(m , w)|[w].\ ] ] let @xmath238 be the completed quantum polynomial algebra @xmath239 $ ] , where the multiplication rule is @xmath240 .
then the map @xmath241 sending @xmath132\to a_m^{-1}x^{{\overline{m}}}$ ] is an algebra morphism @xcite . here
, we use the slanted @xmath242 to distinguish the one with target @xmath243 . apply @xmath3 to both sides , we get @xmath244}[u ] \oldint \sum_{[v]}|{\operatorname{epi}}_q(m , v)|[v ] = \oldint\sum_{[w]}|{\operatorname{hom}}_q(m , w)|[w ] \\
\leftrightarrow & \sum_\gamma r_\gamma(q)x^\gamma\sum_{[v]}a_v^{-1}|{\operatorname{epi}}_q(m , v)|x^{{\overline{v } } } = \sum_{[w]}a_w^{-1}|{\operatorname{hom}}_q(m , w)|x^{{\overline{w } } } \\ \leftrightarrow & \sum_\gamma r_\gamma(q)x^\gamma\sum_\beta |{\operatorname{gr}}^\beta(m)| x^\beta= \sum_{[w]}\frac{|{\mathcal{o}}_w|}{|{\operatorname{gl}}_\alpha|}|{\operatorname{hom}}_q(m , w)|x^\alpha \quad ( \alpha={\overline{w } } )
\\ \leftrightarrow & \sum_{\beta+\gamma=\alpha}{\langle\gamma,\beta\rangle}^{-1}r_{\gamma}(q)|{\operatorname{gr}}^{\beta}(m)|=\sum_{w\in{\operatorname{rep}}_\alpha(q)}\frac{|{\mathcal{o}}_w|}{|{\operatorname{gl}}_\alpha|}|{\operatorname{hom}}_q(m , w)|.\end{aligned}\ ] ] now we set @xmath245 , then the formula follows from lemma [ l : variety ] .
the dual formula can be obtained by applying @xmath3 to the identity : @xmath237}|{\operatorname{mon}}_q(u , m)|[u]\big)\cdot\big(\sum_{[v]}[v]\big ) = \sum_{[v]}|{\operatorname{hom}}(w , m)|[w].\ ] ] [ r : gen ] let @xmath246 be the generating functions @xmath247 , and @xmath248 if we set @xmath249 then lemma [ l : rep ] can be rewritten as @xmath250)=r(q)f^\infty(e ) , \text{\quad and\quad } r(q^\circ[e])=f_\infty(e)r(q).\ ] ] a representation @xmath176 is called _ polynomial - count _
, if all its grassmannians @xmath251 are polynomial - count .
it is called _ add - polynomial - count _
, if each @xmath34 is polynomial - count .
@xmath34 is polynomial - count if and only if @xmath252)$ ] is polynomial - count for any @xmath10 . if @xmath26 is add - polynomial - count , then @xmath27 $ ] is f - polynomial - count by .
it follows from lemma [ l : tao ] , [ l : frep ] , and [ l : rep ] that [ t : ext ] @xmath253)$ ] can be explicitly counted in terms of @xmath254 s .
in particular , if @xmath26 is add - polynomial - count , then each @xmath255)$ ] is polynomial - count when it is a geometric quotient .
we will see in the last section that the assumption of being a geometric quotient can be dropped .
polynomial - count representations of quivers will be studied in detail in @xcite .
this class includes rigid representations because of ( * ? ? ?
* corollary 6.4 ) , but there are many more
. is there a representation , which is polynomial - count but not add - polynomial - count ?
for any representation @xmath26 , the @xmath26-homological stratification of @xmath43 is the decomposition of @xmath43 into ( finite ) disjoint union of locally closed subvarieties @xmath256 , where @xmath257 by lemma [ l : variety ] , we know that for @xmath258 , @xmath259)|=\sum_h |{\operatorname{rep}}_\alpha(q;e , h)|q^{nh}.\ ] ] the coefficient matrix of above linear system is a non - degenerated vandermonde - type matrix , so we can solve all @xmath256 . in particular , @xmath256 is polynomial - count if and only if @xmath26 is add - polynomial - count .
we will see that the above is still true for @xmath260 let @xmath261 be any slope function for @xmath1 .
the ( negative ) extension @xmath262 to @xmath185 $ ] with respect to an dimension vector @xmath10 is @xmath263 , where @xmath264 and @xmath265 for some sufficiently small positive @xmath266 .
similarly , we define the ( positive ) extension of @xmath267 to @xmath268 $ ] with respect to @xmath10 as @xmath269 , where @xmath270 , and @xmath271 .
the following lemma was proved in ( * ? ? ?
* theorem 5.2 ) for @xmath26 projective , but the argument goes through for any @xmath26 .
we have the following identity in @xmath238 : @xmath272)x^\gamma\big ) = \sum_\alpha \big(\sum_{m\in{\operatorname{rep}}_\alpha^\mu(q)}\frac{|{\mathcal{o}}_m||{\operatorname{hom}}_q(ne , m)|}{|{\operatorname{gl}}_{\alpha}||{\operatorname{gl}}_n|}\big)x^\alpha.\ ] ] [ t : hs ] @xmath273 can be explicitly computed from @xmath274 .
when @xmath26 is add - polynomial - count and @xmath275 is a geometric quotient , each homological strata on @xmath275 is polynomial - count . according to theorem [ t : ext ] , all @xmath276)$ ]
s can be computed from @xmath274 , and so does the right hand side of .
notice that @xmath277 we can invert the same vandermonde - type matrix as before to solve @xmath273 .
consider the @xmath223-arrow kronecker quiver @xmath278 and its extension by an @xmath279-dimensional representation @xmath26 .
then we can view the algebra @xmath280 $ ] as an algebra coextended from @xmath281 by a @xmath282-dimensional representation @xmath283 .
it follows from remark [ r : gen ] that [ p : dual ] @xmath284 and @xmath285 are related by @xmath286)=f_\infty(e^\circ)r(k_m).\ ] ] in particular , if @xmath26 is add - polynomial - count , then so is @xmath283 . let @xmath287 $ ] be the algebra coextended from @xmath281 by a representation @xmath26 of dimension @xmath288 . for any dimension vector @xmath289 of @xmath281 ,
there is a unique choice of weight @xmath64 up to scalar such that @xmath290 . for the rest of this section
, we always take @xmath261 for different @xmath10
. the first two isomorphisms below can be easily established by lemma [ l : variety ] .
@xmath291 and @xmath292 .
+ assume that @xmath26 is not too special so that @xmath293 is empty .
@xmath294-[\epsilon_2 - 1])|{\operatorname{gr}}_{(1,1)}(e)|,\\ & |{\operatorname{mod}}_{(2,2,1)}^{\mu_-}(a)|=|{\operatorname{gr}}_{(2,2)}(e)|+([2m-1]-[\epsilon_2 - 1])|{\operatorname{gr}}_{(2,1)}(e)|,\\ & \quad\quad\cdots\cdots\quad \ ] ] where @xmath295 $ ] is the quantum number .
consider the quiver @xmath296 with relation @xmath297 .
the corresponding algebra @xmath6 is one - point - extended from the dynkin quiver @xmath2 by the simple @xmath298 .
so @xmath299 can be computed by lemma [ l : tao ] .
note that @xmath300 is just the usual grassmannian variety @xmath301 .
it follows that the quiver @xmath302 with relations @xmath303 is polynomial - count .
this algebra is extended from the kronecker quiver @xmath304 by a decomposable non - rigid representation of dimension @xmath305 .
[ ex:2n ] fix @xmath306 , we consider the quiver @xmath307 @xmath308 with relation @xmath309 , where @xmath310 and @xmath311 . it is extended form @xmath278 by @xmath312 presented by @xmath313 it is also coextended from @xmath304 by the exceptional @xmath314 presented by @xmath315 where @xmath316 .
although @xmath312 is not rigid , it follows from proposition [ p : dual ] that it is add - polynomial - count .
@xmath317 can be recursively computed by the cluster theory and lemma [ l : gralg ] .
a closed formula was given in ( * ? ? ?
* theorem 4.3 ) .
@xmath318}{\bigl[\begin{smallmatrix}\gamma_2 + 1 \\ \gamma_1\end{smallmatrix}\bigr ] } & \text { otherwise , } \end{cases}\end{aligned}\ ] ] where @xmath319}$ ] is the quantum binomial coefficient .
now we recall ( * ? ? ?
* proposition 2.8 ) .
we also draw some easy consequences , which are useful for counting the grassmannian of representations .
[ l : gralg ] assume that @xmath320 and @xmath321 .
@xmath322}\frac{|{\operatorname{ext}}_q(u , v)_w|}{|{\operatorname{ext}}_q(u , v)|}|{\operatorname{gr}}_{\gamma}(w)|.\ ] ] now suppose that @xmath323 .
then @xmath324 hence , if both @xmath325 and @xmath326 are ( add)-polynomial - count , then so is @xmath327 . moreover , if @xmath328 and @xmath125 is the only non - trivial middle term of the extensions , then @xmath329 we add one arrow to example [ ex:2n ] : @xmath330 then it is extended from @xmath278 by @xmath331 , or coextended from @xmath304 by @xmath332 .
let @xmath333 $ ] .
since @xmath334 , we can use lemma [ l : gralg ] or compute directly @xmath335 .
so we are able to find all @xmath336 . for example ,
@xmath337+[3]-1,\\ & |{\operatorname{mod}}_{(2,2,1)}^{\mu_-}(a_n)|=q^4 + 2q^3 + 4q^2 + 2q + 1 . \ ] ] however , all @xmath338 are different , they are hirzebruch surfaces @xmath339 @xcite
. consider quiver @xmath340 with relation @xmath341 .
it is coextended from @xmath342 by a rigid module presented by @xmath343 .
similar calculation as before gives @xmath344[3],\\ & |{\operatorname{mod}}_{(2,1,1)}^{\mu_\pm}(a)|=|{\operatorname{mod}}_{(1,1,2)}^{\mu_\pm}(a)|=[3],\\ & |{\operatorname{mod}}_{(1,2,1)}^{\mu_\pm}(a)|=[3][5],\\ & |{\operatorname{mod}}_{(2,2,1)}^{\mu_-}(a)|=|{\operatorname{mod}}_{(1,2,2)}^{\mu_-}(a)|=[3][5](1,0,1),\\ & |{\operatorname{mod}}_{(1,2,2)}^{\mu_-}(a)|=|{\operatorname{mod}}_{(2,2,1)}^{\mu_+}(a)|=[3](1,1,3,3,3,1,1).\end{aligned}\ ] ] the first one is @xcite a divisor @xmath345 on @xmath346 of bidegree @xmath347 , or equivalently the complete flag variety @xmath348 of @xmath349 . now consider the deformation @xmath350 of @xmath26 , where @xmath351 .
since @xmath352 with @xmath26 the only non - trivial middle term , we can compute @xmath353 using lemma [ l : gralg ] .
@xmath354x^{(1,0)}+[2]^2x^{(1,1)}+[2]x^{(2,1)}+x^{(0,2)}+[5]x^{(1,2)}+{\bigl[\begin{smallmatrix}5\\2\end{smallmatrix}\bigr]}x^{(2,2)}+\cdots.\ ] ] @xmath355[5],\\ & |{\operatorname{mod}}_{(2,2,1)}^{\mu_-}(a)|=|{\operatorname{mod}}_{(1,2,2)}^{\mu_+}(a)| = [ 3][5](1,0,1),\\ & |{\operatorname{mod}}_{(1,2,2)}^{\mu_-}(a)|=|{\operatorname{mod}}_{(2,2,1)}^{\mu_+}(a)|=[3](1,1,4,4,3,1,1).\end{aligned}\ ] ] note that the first one is irreducible and singular by lemma [ l : polycount ] .
consider quiver @xmath340 with relation @xmath309 , where @xmath356 .
it is coextended from @xmath342 by @xmath26 presented by the following base diagram .
the black dots are a basis in @xmath357 ; while the white dots are a basis in @xmath358 .
the letter on an arrow means the identity map on the arrow of the same letter .
@xmath359 it is known @xcite that for a general representation @xmath360 of dimension @xmath361 , @xmath362 is an elliptic curve .
so @xmath360 is _ not _ polynomial - count .
however , for this special @xmath26 , @xmath363 is three @xmath364 s intersecting at a point . with a little effort one can show that @xmath26 is actually polynomial - count .
let us consider a category , which is universal in the sense that it contains all one - point extensions of @xmath1 as its full subcategories .
it is clearly the module category of @xmath366 , where @xmath2 is the dynkin quiver @xmath367 .
the quiver of @xmath35 is composed of two copies of @xmath1 corresponding to two idempotents of @xmath368 , and _
morphism arrows _ connecting the same vertices in two different copies .
the relations are obviously the commuting diagram relations . by abuse of notation
, we use @xmath365 to denote such a quiver with relations .
so the dimension vector of @xmath365 is composed of two dimension vectors of @xmath1 , say @xmath369 . by convention ,
@xmath10 correspond to the quiver sending morphism arrows .
let @xmath326 be an @xmath10-dimensional @xmath4-vector space .
we denote by @xmath370 the incidence variety @xmath371 and by @xmath372 the incidence variety @xmath373 [ l : a2q ] @xmath374 is a fibre bundle with fibre @xmath375 so @xmath376 is equal to @xmath377 where @xmath378 } { { \langlec - c_e , c_e - c_d\rangle}{\langled - c_d , c_d\rangle}{\langlee - d - c_e+c_d , d+c_e - c_d\rangle}{\langle\gamma_d - c - e+c_e , c+e - c_e\rangle}}$ ] .
we sketch the fibre bundle construction by a picture . after fixing an elements in @xmath379
, we need to fill in the missing part for a @xmath380-dimensional representation of @xmath365 .
similar to lemma [ l : frep ] , the missing part consist of a @xmath381-dimensional representation @xmath19 , a @xmath382-dimensional representation @xmath214 , and a bunch of linear maps from @xmath215 to @xmath216 , as indicated below .
@xmath383 the first step is to choose a configuration of image spaces of the vertical and diagonal morphism arrows .
let @xmath384 be the rank vector of the morphism arrows of @xmath214 and @xmath19 respectively , and @xmath385 be the rank vector of the diagonal morphism arrows .
let @xmath386 be the rank vector of the diagonal morphism arrows restricted on the kernel of the morphism arrows of @xmath214 .
the second step is to stuff in the following order the lower part of @xmath19 , the upper part of @xmath214 , the rest part of @xmath19 and @xmath214 , and other diagonal arrows .
they corresponds to @xmath387 , , , and respectively .
we leave the details to the readers .
to compute @xmath388 , we only need to count the incidence varieties .
we use the transitive action of @xmath45 and count the stabilizers .
the following formulas are immediate . @xmath389 put the fibre bundle structure and these equations together , and we obtain what we desire .
this result can be generalized to the @xmath390-step frep varieties .
so we conclude that the algebra @xmath365 is f - polynomial - count . for the @xmath391-step case
, it suffices to set @xmath392 .
@xmath393 this formula has a dual version : @xmath394 alternatively , this corollary can be proved by a hall algebra method similar to lemma [ l : rep ] .
consider the following identity in the algebra @xmath395 .
@xmath396\otimes\sum_{[u]}[u]\big)\big(\sum_{[m],[v]}|{\operatorname{epi}}_q(m , v)|[m]\otimes [ v]\big ) = \sum_{[m],[w]}|{\operatorname{hom}}_q(m , w)|[m]\otimes [ w].\ ] ] applying the character @xmath397 to the both sides , we see the result immediately .
it follows from lemma [ l : tao ] and [ l : a2q ] that [ t : a2q ] if @xmath398 is a geometric quotient , then it has a counting polynomial , which can be explicitly computed .
we will see in the last section that the assumption of being a geometric quotient is unnecessary .
this result is known ( * ? ? ?
* theorem 4.3 ) for some special choices of @xmath10 and @xmath8 .
consider the 3-arrow kronecker quiver @xmath342 with dimension vectors @xmath399 and @xmath400 .
let @xmath54 be a general representation of dimension @xmath10 , then @xmath54 has no subrepresentation of dimension @xmath401 .
so the projection @xmath402 is an isomorphism .
we can use the algorithm in ( * ? ? ?
* corollary 4.4 ) to find that @xmath403[2]^2(1,4,2,8,5,8,2,4,1),\end{aligned}\ ] ] where @xmath404 is the slope function constructed in ( * ? ? ?
* section 1 ) .
recall that @xmath405 for some sufficiently small @xmath288 .
now we change @xmath406 to @xmath407 , then @xmath408 there is no difficulty to generalize the above results to @xmath409 .
in fact , we will do it more generally in @xcite .
if @xmath26 is add - polynomial - counting , then @xmath27\otimes ka_2 $ ] is f - polynomial counting .
let us come back to general @xmath7 .
consider the map @xmath410=\frac{|{\operatorname{gr}}_\gamma(w)|}{a_w}$ ] as in ( * ? ? ?
* section 2 ) .
if we apply this map to @xmath146 , we get @xmath411 we known from ( * ? ? ?
* lemma 1.2 ) that when @xmath11 is a geometric quotient , this number is equal to @xmath412 for some slope @xmath413 . to compute @xmath414
, we apply @xmath415 to as before .
we define @xmath416 then @xmath417 let @xmath418 where @xmath419 is the projection .
then @xmath420 is stratified by the locally closed subvarieties @xmath421 when @xmath422 , for any decompositions @xmath127 and @xmath423 , @xmath424 is the same as the incidence variety @xmath425 the proof of the following lemma is similar to that of lemma [ l : rep ] and [ l : a2q ] , so we leave it for the readers .
[ l : frep2 ] @xmath426)\to { \operatorname{gr}}_{{\tilde{a}\cap\tilde{\gamma}}}^{\tilde{c}}(\tilde{\alpha})$ ] is a fibre bundle with fibre @xmath427)\times { \operatorname{rep}}_{(b , a_+)}(q^\circ[e])\times { \operatorname{rep}}_{(\gamma - c,\gamma_+)}(q^\circ[e])\times { \operatorname{rep}}_{\tilde{c}}(q^\circ[e])\\ & \times \prod_{a\in q_1}{\operatorname{hom}}(k^{(\beta - b)(ta)},k^{(\gamma+b)(ha)})\times { \operatorname{hom}}(k^{b(ta)},k^{c(ha ) } ) \times { \operatorname{hom}}(k^{(\gamma - c)(ta)},k^{c(ha)}).\end{aligned}\ ] ] so @xmath428)=\sum_{\tilde{b}+\tilde{c}=\tilde{a}}t_{(\beta,\gamma , b , c)}\cdot r_{(\beta - b,\alpha_+)}r_{(b , a_+)}r_{(\gamma - c,\gamma_+)}r_{\tilde{c}},\ ] ] where @xmath429}{\bigl[\begin{smallmatrix}a_+\\b_+\end{smallmatrix}\bigr]}{\bigl[\begin{smallmatrix}\gamma_+\\(\gamma - c)_+\end{smallmatrix}\bigr]}|{\operatorname{gl}}_{(\gamma+b)_+}||{\operatorname{gl}}_{c_+}|}{{\langle\beta - b,\gamma+b\rangle}{\langleb , c\rangle}{\langle\gamma - c , c\rangle}},$ ] and @xmath430)$ ]
. readers can easily write out the formula for the dual case @xmath185 $ ] .
this lemma can be recursively generalized to the @xmath390-step case : @xmath431 .
[ t : a2ext ] if @xmath26 is add - polynomial - count and @xmath255)$ ] is a geometric quotient , then @xmath432)}|{\operatorname{gr}}_\gamma(m)|$ ] is polynomial - count for any @xmath433 . as in (
* section ) , we can also consider the @xmath434-step analog of @xmath435 : @xmath436={a_w}^{-1}|{\operatorname{fl}}_{\gamma_t\cdots\gamma_1}(w)|.\ ] ] everything can be generalized to this case without any essential difficulty .
finally we consider the map @xmath21 from the hall algebra @xmath131 to the formal power series algebra
@xmath437 $ ] in @xmath438 variables as in ( * ? ? ?
* section 8) : @xmath439=a_w^{-1}\sum_{i=0 } ( -1)^{i+1 } f_i(w)x^\alpha,\ ] ] where @xmath440 is the number of @xmath90-step filtrations of @xmath125 .
we recall from ( * ? ? ?
* lemma 8.3 ) that the number @xmath441 has a neat formula in terms of the multiplicities of simple summands of @xmath125 .
fix a slope function @xmath8 and a slope @xmath442 .
let @xmath76 be the abelian subcategory of all semistable representations with slope @xmath75 , and @xmath443 $ ] .
let us denote by @xmath444 the number of @xmath10-dimensional absolutely stable representations and @xmath445 .
the absolute ( resp .
relative ) poincar series of @xmath446 at @xmath75 is @xmath447 ( resp .
@xmath448 ) .
our convention is that the relative ones have constant term @xmath391 , but @xmath449 for the absolute ones .
it was proved in ( * ? ? ?
* theorem 4.1 ) that @xmath450 where @xmath451 is the plethystic exponential in the @xmath452-ring @xmath437 $ ] ( * ? ? ?
* section 2 ) .
moreover , it is known ( * ? ? ?
* theorem 8.3 ) that @xmath453 actually , for both the argument would work for any algebra not necessarily hereditary . to compute @xmath454
, we apply @xmath21 to each individual @xmath455 for @xmath456 using .
it follows from lemma [ l : frep2 ] and its @xmath434-step generalization that the series @xmath457 has all coefficients polynomials in @xmath5 , and so are @xmath458 .
99 i. assem , d. simson , a. skowroski , _ elements of the representation theory of associative algebras , _ london mathematical society student texts 65 , cambridge university press , 2006 . w. crawley - boevey , m. van den bergh , _ absolutely indecomposable representations and kac - moody lie algebras , _ invent .
( 2004 ) , no .
3 , 537559 . h. derksen , j. weyman , _ the combinatorics of quiver representation , _ ann .
fourier ( grenoble ) 61 ( 2011 ) , no .
3 , 10611131 .
j. engel , m. reineke , _ smooth models of quiver moduli , _ math .
z. 262 ( 2009 ) , no .
4 , 817848 . j. fei , _ counting using hall algebras i. quivers , _ j. algebra 372 ( 2012 ) , 542559 . _
counting using hall algebras iv .
tensor products of quivers , _ in preparation . _ moduli and tilting ii .
extensions from quivers , _ preprint . _ polynomial - count representations of quivers , _ in preparation
king , _ moduli of representations of finite - dimensional algebras , _ quart .
oxford ser .
( 2 ) 45 ( 1994 ) , no .
180 , 515530 .
s. mozgovoy , m. reineke , _ on the number of stable quiver representations over finite fields , _ j. pure appl .
algebra 213 ( 2009 ) , no .
4 , 430439 .
m. reineke , _ the harder - narasimhan system in quantum groups and cohomology of quiver moduli , _ invent .
( 2003 ) , no .
2 , 349368 . , _ counting rational points of quiver moduli , _ int . math
. res . not .
2006 , art.id 70456 , 19 pp .
, _ moduli of representations of quivers , _ trends in representation theory of algebras and related topics , 589637 , ems ser .
soc . , zrich , 2008 . o. schiffmann , _ lectures on hall algebras , _ preprint , arxiv:0611617 .
c. sznt _ on the cardinalities of kronecker quiver grassmannians , _ math . z. 269 ( 2011 ) , no . 3 - 4 , 833846 .
j. xiao , _ drinfeld double and ringel - green theory of hall algebras , _ j. algebra 190 , no .
1 , 100144 ( 1997 ) . | we count the @xmath0-rational points of git quotients of quiver representations with relations .
we focus on two types of algebras
one is one - point extended from a quiver @xmath1 , and the other is the dynkin @xmath2 tensored with @xmath1 . for both
, we obtain explicit formulas .
we study when they are polynomial - count .
we follow the similar line as in the first paper but algebraic manipulations in hall algebra will be replaced by corresponding geometric constructions .
[ multiblock footnote omitted ] |
broadband observations of gamma - ray burst afterglows can in principle be used to constrain fundamental physical parameters of the explosion . in the fireball model
, a relativistic blast wave expands into the surrounding medium , its hydrodymanical evolution being strongly influenced by the density structure of the medium as well as by the energy content and geometry ( in particular collimation ) of the ejecta .
the temporal behavior of the afterglow emission which arises from the shocked gas depends on the shock evolution , and the partition of energy between the magnetic field and relativistic electrons , and can therefore probe these physical parameters given data of sufficient quality . in this paper , we report the synthesized results from our multi - frequency followup campaign on the relatively bright grb 000926 .
this campaign was aimed at studying the evolution of the afterglow to constrain the model parameters described above .
@xcite have reported the results from our multi - band ( @xmath2 ) optical monitoring .
we combine these data with 4 epochs taken with the _ hubble space telescope _
wfpc2 , with _ chandra x - ray observatory _ ( cxo ) target of opportunity ( too ) observations , and with multi - frequency radio monitoring from the very large array ( vla ) , the ryle telescope , and the owens valley radio observatory .
we interpret the resulting broadband lightcurve in the context of a theoretical afterglow model .
lclc oct 7.25 & f450 & 2x1100 ( 1 orbit ) & 24.98 @xmath3 0.07 + oct 7.35 & f606 & 4x1100 ( 2 orbits ) & 24.54 @xmath3 0.03 + oct 7.49 & f814 & 4x1100 ( 2 orbits ) & 23.89 @xmath3 0.03 + oct 16.08 & f450 & 2x1100 ( 1 orbit ) & 25.82 @xmath3 0.14 + oct 16.18 & f606 & 4x1100 ( 2 orbits ) & 24.27 @xmath3 0.03 + oct 16.32 & f814 & 4x1100 ( 2 orbits ) & 24.87 @xmath3 0.05 + oct 25.05 & f450 & 2x1100 ( 1 orbit ) & 25.59 @xmath3 0.12 + oct 25.21 & f606 & 4x1100 ( 2 orbits ) & 25.45 @xmath3 0.03 + oct 25.35 & f814 & 4x1100 ( 2 orbits ) & 24.96 @xmath3 0.05 + dec 16.02 & f606 & 6x1000 ( 3 orbits ) & 25.58 @xmath3 0.03 + dec 16.90 & f814 & 4x1100 ( 2 orbits ) & 25.24 @xmath3 0.07 + cccc [ tab - hst ] 10.3570 & @xmath4 ( b ) & @xmath5 + 19.1870 & @xmath4 ( b ) & @xmath6 + 28.2170 & @xmath4 ( b ) & @xmath7 + 10.3570 & @xmath8 ( v ) & @xmath9 + 19.1870 & @xmath8 ( v ) & @xmath10 + 28.2170 & @xmath8 ( v ) & @xmath11 + 80.4670 & @xmath8 ( v ) & @xmath12 + 10.3570 & @xmath13 ( r ) & @xmath14 + 19.1870 & @xmath13 ( r ) & @xmath15 + 28.2170 & @xmath13 ( r ) & @xmath16 + 80.4670 & @xmath13 ( r ) & @xmath17 + 10.3570 & @xmath18 ( i ) & @xmath19 + 19.1870 & @xmath18 ( i ) & @xmath20 + 28.2170 & @xmath18 ( i ) & @xmath21 + 80.4670 & @xmath18 ( i ) & @xmath22 + cccc [ tab - xray2 ] 2.7700 & @xmath23 & @xmath24 + 13.477 & @xmath23 & @xmath25 + 2.2780 & @xmath26 & @xmath27 + 2.7700 & @xmath26 & @xmath28 + 13.477 & @xmath26 & @xmath29 + ccccc [ tab - xray ] sax & sep .
29.03 29.53 & 1.5 8 & @xmath30 + cxo - p1 & sep.29.674 29.851 & 0.2 1.5 & @xmath31 + cxo - p1 & sep.29.674 29.851 & 1.5 8 & @xmath32 + cxo - p2 & oct.10.176 10.760 & 0.2 1.5 & @xmath33 + cxo - p2 & oct.10.176 10.760 & 1.5 8 & @xmath34 + lcc [ tab - radio ] sep 29.708 & 98.48 & @xmath35 + oct 1.708 & 98.48 & @xmath36 + oct 4.186 & 22.5 & @xmath37 + oct 5.216 & 22.5 & @xmath38 + oct 16.721 & 22.5 & @xmath39 + sep 28.81 & 15.0 & @xmath40 + sep 30.91 & 15.0 & @xmath41 + oct 1.69 & 15.0 & @xmath42 + oct 5.751 & 15.0 & @xmath43 + oct 11.751 & 15.0 & @xmath44 + sep 28.17 & 8.46 & @xmath45 + sep 28.97 & 8.46 & @xmath46 + sep 29.726 & 8.46 & @xmath47 + oct 4.186 & 8.46 & @xmath48 + oct 5.216 & 8.46 & @xmath49 + oct 7.771 & 8.46 & @xmath50 + oct 8.291 & 8.46 & @xmath51 + oct 10.281 & 8.46 & @xmath52 + oct 12.771 & 8.46 & @xmath53 + oct 15.681 & 8.46 & @xmath54 + oct 23.161 & 8.46 & @xmath55 + oct 27.131 & 8.46 & @xmath56 + oct 30.181 & 8.46 & @xmath57 + nov 26.64 & 8.46 & @xmath58
+ dec 18.95 & 8.46 & @xmath59 + jan 29.44 & 8.46 & @xmath60 + feb 9.73 & 8.46 & @xmath61 + sep 28.17 & 4.86 & @xmath62 + sep 28.97 & 4.86 & @xmath63 + sep 29.726 & 4.86 & @xmath64 + oct 4.186 & 4.86 & @xmath65 + oct 7.741 & 4.86 & @xmath66 + oct 8.701 & 4.86 & @xmath67 + oct 30.201 & 4.86 & @xmath68 + nov 18.03 & 4.86 & @xmath69 + jan 6.53 & 4.86 & @xmath70 + feb 2.47 & 4.86 & @xmath71 + feb 19.28 & 4.86 & @xmath72 + dec 16.58 & 1.43 & @xmath73 + the interplanetary network discovered grb 000926 on 2000 sep 26.993 ut @xcite .
the afterglow of this 25 s long event was identified less than a day later @xcite .
the redshift , measured from optical absorption features , is @xmath74 @xcite .
the afterglow was well - monitored in the optical @xcite , and was detected in the ir @xcite .
here we describe _ hubble space telescope ( hst ) , chandra x - ray observatory ( cxo ) _ and radio observations . as part of an _ hst _ cycle 9 program we observed grb 000926 at four epochs with the wide field planetary camera 2 ( wfpc2 ) , with the ot placed on wfpc ccd#3 .
in the first three epochs we observed at three passbands corresponding to the f450w , f606w , and f814w filters , and in the final epoch we used only the f606w and f814w filters .
these observations took place between oct 7.25 ( 10.26 days after the grb ) and dec 16.9 ( 81.9 days after ) .
table 1 shows a log of the hst observations , along with the magnitude derived for each filter for the 2-pixel radius region surrounding the ot .
we determined the aperture correction using a 2 5 pixel radius , and we quote the corresponding 5 pixel radius magnitude .
we calibrated the zeropoints , and converted the wfpc2 filters to johnson cousins magnitudes using the color transformations from @xcite .
we estimate the associated calibration uncertainty to be about 0.10 magnitude in @xmath75 and @xmath76 bands , and 0.20 magnitudes in the @xmath77 band . to perform the transformations
, we have interpolated the 3-filter wfpc2 data to obtain photometric points in @xmath78 and @xmath77 .
measurements in only three of the bands are , therefore , truly independent .
we tied the wfpc2 calibration to the ground - based data of @xcite using their tertiary star at position ra @xmath79 dec @xmath80 ( j2000 ) .
the magnitudes we derive for this star agree with those obtained by @xcite to within 0.10 mag in @xmath81 , and @xmath76 , and to within @xmath82 mag in @xmath77 . @xcite
provide further details on the image reduction and magnitude calibration . from flattening in the late - time ( @xmath83 days ) optical lightcurve
we infer that the hst data are contaminated by a contribution from the host galaxy .
we do not actually resolve the host in our hst images , and we observe flattening even with our small ( 0.14@xmath84 ) extraction radius .
this implies that the host itself is compact , or that the constant emission is from a bright knot or filament associated with the host .
future hst observations scheduled for late may 2001 may resolve this feature , and we will present these results elsewhere . to determine the flux from the ot we subtract this constant contribution . to derive the magnitude for the constant term , we fit the ground - based optical lightcurve from price _
et al_. ( 2001 ) combined with our hst data to a decaying function with a constant added : @xmath85 the decaying function , taken from @xcite , has no physical significance , but provides a simple and general parametric description of the data . here
@xmath86 and @xmath87 are the early and late time asymptotic temporal slopes respectively , @xmath88 is the time of the temporal slope break , @xmath89 is a parameter that determines the sharpness of the transition , and @xmath90 is the spectral slope . from the combined ground - based and hst data we derive @xmath91,@xmath92,@xmath93 days , @xmath94 .
all errors in this paper are 1-@xmath95 .
the best - fit value for the constant is @xmath96 , @xmath97 , @xmath98 , @xmath99 .
we note that our fit values are similar to those derived by price et al .
( 2001 ) , but the late - time , high - resolution hst data allow us to determine the flux from the compact region near the ot . we derive the ot magnitudes by subtracting the constant term , @xmath100 , given above from each band , correcting for the foreground galactic extinction of @xmath101 mag @xcite .
table 2 contains the resulting dereddened , host - subtracted ot flux values converted to @xmath102jy .
the measurement errors include a contribution ( 1@xmath103 ) of 0.026(b ) , 0.016 ( v ) , 0.018 ( r ) , 0.020 ( i ) @xmath102jy added in quadrature to the statistical and systematic calibration errors to reflect the uncertainty in the constant host term derived from the fit .
the _ bepposax _ medium energy concentrating spectrometer ( mecs ) instrument ( sensitive from 1.6 10 kev ) discovered the x - ray afterglow of grb 000926 in an observation made on sep 29.03
29.52 @xcite .
@xcite observed the source for 10 ksec as part of a cycle 1 cxo program using the acis s3 backside - illuminated chip on sep 29.674 29.851 ( referred to here as cxo - p1 ) .
supporting our hst cycle 9 program , cxo again observed grb 000926 in a 33-ksec long too taken oct 10.17610.76 ut ( cxo - p2 ) , also with the transient placed on the acis s3 backside - illuminated chip .
the afterglow was clearly detected in each of these observations .
the transient was not detected in the _ bepposax _ low energy concentrating spectrometer ( co - aligned with the mecs , and sensitive in the 0.1 10 kev band ) , however the effective exposure time was only 5 ksec .
@xcite present a detailed spectral analysis of the x - ray data , including combined fits to all pointings .
we analyzed the data from cxo - p1 and cxo - p2 using software provided by the _
chandra _ x - ray observatory center ( cxc ) to filter the events , extract source counts and subtract background . for the latter we used an annular region surrounding the source .
we fit the two _ chandra _ pointings separately using a powerlaw model plus absorption . the best fit photon spectral index for the cxo - p1 data
is @xmath104 and for cxo - p2 is @xmath105 , with best - fit hydrogen column depth @xmath106 @xmath107 ( cxo - p1 ) and @xmath108 @xmath107 ( cxo - p2 ) .
we find no evidence for additional absorption in the x - ray spectrum at the redshift of the host ( @xmath109 ) .
this is consistent with the host @xmath110 of 0.1 mag derived from model fits to the optical data ( see below ) , if we adopt typical dust to gas ratios .
for an lmc - like extinction curve , for example , this @xmath110 corresponds to column depth of @xmath111 @xmath107 for a host at @xmath109 .
given the host redshift , such a low column is not detectable in the x - ray spectrum , and so no correction for absorption in the grb host is necessary .
the detection statistics in the x - ray are limited , so for the purposes of our modeling we divided the energy range into soft ( 0.2 1.5 kev ) and hard ( 1.5 8 kev ) bands .
we converted counts to flux using exposure maps weighted in energy using a photon powerlaw spectral index of @xmath112 and @xmath113 @xmath107 , corresponding to the value for our own galaxy , as determined by w3nh .
we adopt this value since it is consistent with that derived from our spectral modeling , and we find no significant evidence for additional absoprtion in the host . for the _ sax _
observation , we used the best fit spectral model to determine the hard band flux ( the mecs response does not allow a soft - band flux to be determined ) .
table 3 shows the flux values , not corrected for galactic absorption , for the three observations . to determine the center energy for the band
, we took the mean , weighted using a photon index of @xmath112 .
we determined flux errors by adding in quadrature the statistical error and the error due to the uncertainty in spectral slope .
since the flux is decaying during each observation interval , we weight the time of the observation by @xmath114 and average to determine the mean epoch . table 4 lists the center frequencies and flux values for the x - ray transient ( converted to @xmath115jy ) , where we have corrected for galactic absorption .
we obtained radio observations at frequencies from 1.43 to 100 ghz at a number of facilities : the owens valley radio observatory interferometer ( ovro ) , the ryle telescope , and the vla .
table 5 summarizes these observations , organized by frequency .
the 98.48 ghz data were taken at ovro , the 15 ghz data at ryle , and all other frequencies were observed using the vla . in performing the observations , reducing the data and deriving flux errors we adopted the methodology described in detail in @xcite , @xcite , and @xcite . for the ovro observations
, we used the quasar 1726 + 455 for phase calibration and uranus for absolute flux calibration .
to interpret the broadband lightcurve we employ a model for the emission from the relativistic shock ( resulting from the grb explosion ) as it expands into the surrounding medium .
our modeling is quite comprehensive .
it allows for either a constant density medium ( which we refer to as ism ) , or for a density depending on radius as @xmath116 referred to as the wind model , since this profile would be produced by a stellar progenitor s wind @xcite .
we also allow for either isotropic or collimated ejecta , and we fit for extinction in the host galaxy .
we include the effects of inverse compton scattering on the evolution of the synchrotron spectrum , as well as the contribution of ic emission to the observed spectrum .
ic scattering is not included in most models used to fit grb afterglows , but if the ambient medium is of sufficiently high density , it becomes an important effect .
compton emission can dominate the total afterglow cooling rate for months after the event , and can even be directly observed in the x - ray @xcite .
we calculate the shock emission at a given time and frequency from a number of fundamental parameters : the initial isotropic - equivalent energy , @xmath117 in the shock , the electron powerlaw slope , @xmath118 , the electron and magnetic field energy fractions , @xmath119 and @xmath120 , and the density of the circumburst medium .
we employ the equations given by @xcite and @xcite to determine the evolution of the synchrotron spectrum for the constant density medium , and the equations given by @xcite for the wind model .
we include the effect of inverse compton scattering using the treatment given by @xcite .
we calculate the time at which collimation becomes evident ( @xmath121 , where @xmath122 is the lorentz factor of the ejecta ) and the subsequent spectral evolution using @xcite . for the host extinction
we use the parameterization of a(v ) of cardelli et .
al . ( 1989 ) and fitzpatrick & massa ( 1998 ) , joined using the formula from reichart ( 1999 ) in the rest - frame optical / uv , and absorption cross - sections given by @xcite in the rest - frame x - ray .
we consider both lmc or smc - like extinction laws . to fit the broadband lightcurve we vary the fundamental parameters , iterating to minimize @xmath123 .
at any given time the synchrotron emission , which dominates at most energies , has a spectrum characterized by a number of breaks : a low - frequency roll over due to self - absorption at frequency @xmath124 ; a peak at frequency @xmath125 due to the peak energy in the electron distribution ; and a cooling break at @xmath126 , where radiative energy losses become a significant fraction of the electron energy .
these breaks evolve in time as the shock evolves , producing frequency - dependent breaks in the lightcurve . for the case of collimated ejecta
, the lightcurve will begin to steepen when @xmath127 is smaller than the angle into which the emission is beamed due to relativistic effects ( @xmath128 , where @xmath122 is the lorentz factor ) .
we use a sharp transition for this `` jet break '' , and employ spherical evolution laws up until the time of the break . after the jet transition we assume the ejecta expand laterally , and we employ the appropriate asymptotic formulae until the time the ejecta become non - relativistic . the transition to the non - relativistic evolution
is again treated as sharp .
the inverse compton component is also characterized by a series of spectral breaks . with the synchrotron spectrum as the photon source ,
sari & esin ( 2001 ) find that the upscattered component is characterized by three breaks at @xmath129 , @xmath130 , @xmath131 where @xmath132 is the characteristic lorentz factor of the electrons emitting with peak frequency @xmath125 .
we adopt a spectral shape for the ic component similar to the synchrotron spectrum , but with the breaks given by the values in sari & esin ( 2001 ) .
this provides a good approximation to the ic component spectral shape except for at @xmath133 , but for our data this is not a concern ( see below ) . between @xmath134 and @xmath135
there is only a logarithmic correction , which we ignore .
we fit the broadband grb 000926 lightcurve employing the afterglow model described above .
in addition to the data in tables 1 4 , we included the ground - based optical points given in price et.al . ( 2001 ) , as well as @xmath136 band data from @xcite and @xmath137 @xmath138 and @xmath139-band points from @xcite .
we converted the @xmath138 points to @xmath139 using the prescription in @xcite .
all the data are corrected for absorption in our galaxy , and in the case of the optical data we subtracted the contribution from the host .
the uncertainty in the radio flux is dominated by interstellar scintillation over the timescales of interest here . in calculating the @xmath123 for the fit ,
we add an uncertainty to the radio fluxes based on the scintillation envelopes calculated using the fireball size derived from the model .
we have excluded the first 8.46 data point from the fits , as it is likely associated with a separate component from the reverse shock ( see below ) .
we considered several possible cases in fitting the data ; constant density as well as a density gradient for the surrounding medium , and two cases for the ic emission .
the need to consider two cases arises from the fact that for the same sychrotron emission spectrum , sari & esin ( 2001 ) find two possible ic solutions .
these correspond to the two limits @xmath140 , and @xmath141 , where @xmath119 and @xmath120 are the fractions of the total explosion energy that go into accelerating shocked electrons and amplifying the post - shock magnetic field respectively . here
@xmath142 is the fraction of electron energy radiated away , so describes whether cooling is dominant . if @xmath143 ( low ic ) , the ic cooling rate is unimportant compared to that of synchrotron , whereas if @xmath144 ( high ic ) inverse compton cooling dominates the total emission .
table 6 summarizes the physical parameters corresponding to the best fit for each model .
all models we considered required the ejecta to be collimated in order to explain the break in the optical lightcurve at 12 days . in the optical ,
this break is roughly independent of frequency .
extinction of @xmath145 mag , presumably due to the host galaxy , is required to fit the optical spectrum in three out of four cases .
we employed an lmc - like curve for the models , however the magnitude of the extinction is relatively insensitive to the extinction law .
@xcite derived similar values for collimation and extinction based on fits to the optical data alone .
the model which best represents the broadband data is a constant density ism with the high - ic solution .
figures [ fig1]a and [ fig1]b show the broadband spectrum around day 2 and day 10 , with the best - fit model overplotted .
the dot - dashed lines show the separate contributions from the synchrotron and ic components .
synchrotron emission dominates at optical frequencies and below , however on both day 2 and day 10 the model predicts an equal or larger contribution to the x - ray data from ic scattering . from figure 1
it is evident that the approximation to the ic contribution made by assuming the same shape as the synchrotron is not an issue in fitting the data . as mentioned previously , the error in this approximation is only significant below @xmath146 , and for the parameters we derive the synchrotron dominates the total flux in this region by several orders of magnitude .
figures 2 , 3 , and 4 show the x - ray , optical and radio lightcurves .
the @xmath147 for the fit is 124 for 114 degrees of freedom .
the first column of data in table 6 lists the physical parameters for the best fit high - ic solution .
the derived parameters are all reasonable , although we note that @xmath119 of @xmath148 is only marginally consistent with the assumption of adiabatic evolution inherent to our model .
while this will not alter our basic conclusions , in future work we will expand the model to include partially radiative blast waves ( see @xcite and @xcite for treatments of the afterglow evolution in this case ) .
the isotropic - equivalent energy in the blast wave ( @xmath149 in table 6 ) of @xmath150 erg is valid for @xmath151 day .
we use an adiabatic model to fit the data , which assumes the total energy is constant in time . while this is approximately valid at @xmath152 days ( i.e. for the majority of our data ) , the energy we derive will not represent the total _ initial _ blastwave energy , since significant energy is lost to radiation at early times .
in particular , for the high - ic case , the shock wave radiates a significant amount of energy early on , emitted as mev gamma - rays . using the equations in cohen ,
piran & sari ( 1998 ) we estimate the energy in the initial fireball to be about five times higher than that associated with the adiabatically expanding blast wave at a few days , or @xmath153 erg . the observed isotropic energy release in the gamma - ray burst itself was @xmath154 erg @xcite .
recent theoretical work suggests the internal shock process producing the grb should radiate less energy than the blast wave , except under extreme physical conditions @xcite .
the value we derive for the fireball energy is , therefore , reasonable . to fit the low - ic ism model , we initially allowed all of the physical parameters to vary without limitation . for the low - ic case ,
the observed emission at all frequencies is from synchrotron radiation alone .
the fitting program adjusts the physical parameters in order to reproduce all data with this component .
reproducing the broadband observations requires a large fraction of energy in the post - shock magnetic field .
the best - fit value of @xmath155 is unphysical , and so we have fixed @xmath120 at unity the maximum value physically possible .
for either value of @xmath120 the low - ic model provides a poor fit to the data .
the @xmath123 is 187 for 114 datapoints for @xmath156 and 178 for 114 datapoints for @xmath157 .
the largest contribution to the @xmath123 comes from the optical data , where the syncrotron component does not fit the data as well as that derived for the high - ic case .
ccccc [ tab - model ] @xmath123 for 114 data pts & 124 & 187 & 167 & 244 + @xmath158 ( days ) &
@xmath159 & @xmath160 & @xmath161 & @xmath162 + @xmath163 ( days ) & 70@xmath164 & @xmath165 & @xmath166 & @xmath167 + @xmath168 ( @xmath169 erg ) & @xmath170 & @xmath171 & @xmath172 & @xmath173 + @xmath174(ism)/@xmath175(wind ) & 27@xmath176 @xmath1 & 1.6@xmath177 @xmath1 & 3.5@xmath178 & 0.26@xmath179 + @xmath118 & 2.43@xmath180 & 2.20@xmath181 & 3.08@xmath182 & 2.25@xmath182 + @xmath119 ( fraction of e ) & 0.30@xmath183 & 0.16@xmath184 & 0.16@xmath185 & 0.018@xmath186 + @xmath120 ( fraction of e ) & 0.008@xmath186 & 1.0 & @xmath187 & 1.0 + @xmath188 & 0.137@xmath189 & 0.099@xmath186 & 0.103@xmath190 & .047@xmath186 + host a(v ) & 0.12@xmath191 & 0.20@xmath191 & 0.00@xmath185 & 0.20@xmath184 + table 6 provides the physical parameters associated with the best - fit low - ic ism model for @xmath120 fixed at unity .
in addition to the large @xmath120 , the isotropic - equivalent energy in the blast wave of @xmath192 erg for the low - ic case is smaller than the observed gamma - ray energy release .
radiative corrections for this model are much smaller than for the high - ic case , so that the adiabatic value more closely represents the energy in the initial fireball . as pointed out above , @xmath193 is hard to accomodate except under extreme conditions .
figure 2 shows the x - ray data with both high and low - ic ism models overplotted . the high - ic case provides a somewhat better fit to the data , as the low - ic case systematically underpredicts the measurements .
the x - ray observations alone , however , are not sufficient in this case to indicate the presence of an inverse compton component .
this requires the results from the broadband fit .
our calculations show that the ic scattering hardens the x - ray spectrum over some time intervals compared to the prediction low - ic model .
there is only marginal evidence for this in these data ; the observed x - ray spectral photon index is @xmath194 compared with the @xmath195 associated with the synchrotron component .
this emphasizes the role of future , higher - significance x - ray observations in confirming the role if ic scattering in grb afterglows .
we also investigated a wind density profile in both high- and low - ic regimes .
neither of these cases reproduce the data as well as the high - ic ism model .
the low - ic wind solution has a poor @xmath123 of 244 for 114 degrees of freedom .
again we have fixed @xmath196 , since the best - fit value of 174 is unphysical .
the high - ic wind case also has a large @xmath123 of 167 for 114 d.o.f , and under - predicts the 5 ghz radio data by a factor 2 3 .
it does , however , reproduce the x - ray data reasonably well .
we note that early - time optical or high - frequency radio data is particularly important in ruling out the wind model .
this is because the lightcurve evolution after the jet break is similar for ism and wind . directly observing the cooling break evolution provides the greatest leverage , since this evolution is significantly different for wind and ism models .
generally , this requires optical observations or high - frequency radio observations at early epochs ( minutes to hours after the event ) .
unfortunately optical coverage of this event began rather late ( @xmath197 days ) , but future early - time observations of other events will provide a more sensitive probe for density structure in the surrounding medium .
finally , we point out that even for our best fit ( high - ic ism ) solution , the first 8 ghz data point exceeds the model s expectation by a factor of 3.3 ( 7.9 @xmath95 ) .
this discrepancy is worse for the alternative models .
scintillation gains this large are extremely improbable , and it is therefore likely that we are seeing evidence for an additional component due to the reverse shock emission , as was observed in early radio observations of grb 990123 @xcite .
piro _ et al . _
( 2001 ) analyzed the x - ray and optical lightcurves of grb 000926 , and noted that the spectral and temporal properties can not be explained using a standard jet model with moderate ism density . by invoking inverse compton cooling
, we were able to fit the data with a moderate density medium and relatively standard fireball parameters . in order to describe the x - ray and optical lightcurves , piro _ et al .
_ invoke moderate collimation of the fireball ( opening angle @xmath198 ) , and a dense ( @xmath199 @xmath1 ) medium .
this high density results in rapid deceleration of the fireball , and an early transition to the non - relativistic regime .
these features seemingly reproduce the general characteristics of the high - frequency observations .
the high density found by piro _
_ would suggest the event took place in a molecular cloud , and further the low extinction measured in the optical would indicate that the grb destroyed much of the dust in its vicinity .
the radio data provide an important means of distinguishing these two interpretations .
if the fireball expands in a very dense medium , the ejecta decelerate much more rapidly , and at a given time the fireball radius is significantly smaller than for the moderate densities derived in this paper .
the self - absorption frequency , easily observed as a cutoff in the radio spectrum , is correspondingly much higher at a given time for the high - density case .
we have fit a model with the ism density fixed at @xmath200 @xmath1 , and the jet opening angle of @xmath198 to the broadband data set .
we investigated both high - ic and low - ic constant - density solutions , and find we obtain the best fit using the high - ic model .
figure 5 shows the broadband spectrum on day 2 , with a comparison of our best fit with the data .
the radio data quite clearly exclude the high - density possibility , since the self - absorption frequency is more than a decade higher than we measure . on day two ,
the high density model predicts an 8 ghz radio flux of @xmath201jy , while we measure a flux of 200@xmath115jy .
this points out the importance of broadband coverage , including radio , in constraining models for the grb environment .
grb 000926 has one of the highest - quality broadband afterglow lightcurves studied to - date .
we can explain all the general features using a model of a relativistic shock produced by ejecta collimated to a @xmath202 opening angle .
if we correct the isotropic - equivalent gamma - ray energy release for this jet angle we get @xmath203 erg .
this is reasonably accounted for in most currently - popular progenitor models .
our observations require extinction in the host galaxy , with a corresponding hydrogen column of @xmath204 @xmath107 if the dust to gas ratio is typical of the lmc or smc .
this is consistent with the event taking place in a galactic disk .
we find that an explosion occurring in a spatially homogeneous medium best describes the broadband lightcurves , provided that the ratio @xmath205 .
this condition implies that compton emission dominates the total cooling rate , and that ic scattering is potentially directly observable in the afterglow spectrum .
we do , in fact , find that for our best - fit model ic emission dominates over the synchrotron contribution at 1 kev on the 2 10 day timescales spanned by our x - ray data .
detecting the ic emission in the x - ray band directly implies a lower - limit on the density @xcite , in this case @xmath206 @xmath1 , and our best fit yields @xmath207 @xmath1 .
this value for @xmath174 is higher than the average ism density in a typical galaxy , and is consistent with a diffuse interstellar cloud , such as those commonly found in star forming regions .
combining the density with the host @xmath208 derived from our optical observations ( again assuming typical dust to gas ratios ) implies a scale size of 2 pc for the cloud . we wish to thank harvey tananbaum and the cxo operations staff for facilitating the _ chandra _ too observations .
we thank steve beckwith and the hst operations staff for facilitating the wfpc2 observations .
fah acknowledges support from a presidential early career award .
srk and sgd thank nsf for support of their ground - based grb programs .
rs is grateful for support from a nasa atp grant .
rs & tjg acknowledge support from the sherman fairchild foundation . | grb 000926 has one of the best - studied afterglows to - date , with multiple x - ray observations , as well as extensive multi - frequency optical and radio coverage .
broadband afterglow observations , spanning from x - ray to radio frequencies , provide a probe of the density structure of the circumburst medium , as well as of the ejecta energetics , geometry , and the physical parameters of the relativistic blastwave resulting from the explosion .
we present an analysis of _ chandra x - ray observatory _ observations of this event , along with _
hubble space telescope _ and radio monitoring .
we combine these data with ground - based optical and ir observations and fit the synthesized afterglow lightcurve using models where collimated ejecta expand into a surrounding medium .
we find that we can explain the broadband lightcurve with reasonable physical parameters if the cooling is dominated by inverse compton scattering .
for this model , an excess due to inverse compton scattering appears above the best - fit synchrotron spectrum in the x - ray band .
no previous bursts have exhibited this component , and its observation would imply that the grb exploded in a moderately dense ( @xmath0 @xmath1 ) medium , consistent with a diffuse interstellar cloud environment . |
in the last decade , quantile regression has attracted considerable attention .
there are two major reasons for such popularity .
the first is that quantile regression estimation @xcite can be robust to non - gaussian or heavy - tailed data .
in addition , it includes the commonly used least absolute deviation ( lad ) method as a special case .
the second is that the quantile regression model allows practitioners to provide more easily interpretable regression estimates obtained via various quantiles @xmath0 $ ] .
more references about quantile regression estimations and interpretations can be found in the seminal book of @xcite .
further extension of quantile regression to various model and data structures have been found in the literature , e.g. , @xcite for count data , @xcite for power transformed data , @xcite and @xcite for survival analysis , @xcite and @xcite for regression with measurement errors , @xcite for regression with augmented factors , and @xcite for semiparametric varying - coefficient partially linear models , among others .
in addition to the regression context , the quantile technique has been employed to the field of time series ; see , for example , @xcite and @xcite for autoregressive ( ar ) models , @xcite for unstable ar models , and @xcite for generalized autoregressive conditional heteroscedastic ( garch ) models .
it is noteworthy that @xcite established important statistical properties for quantile autoregressive ( qar ) models , and suggested a modified bayesian information criterion ( bic ) to select the order of qar models .
their findings have expanded the classical ar model into a new era , which motivates us to extend the classical box - jenkins approach ( i.e. , model identification , model parameter estimation , and model diagnostics ) from ar to qar models . in the classical ar model
, it is known that model identification usually relies on the partial autocorrelation function ( pacf ) of the observed time series , while model diagnosis commonly depends on the autocorrelation function ( acf ) of model residuals .
detailed illustrations of model identification and diagnosis can be found in @xcite .
the aim of this paper is to introduce two novel measures to examine the linear and partial linear relationships between any two random variables for the given quantile @xmath0 $ ] .
we name them quantile correlation ( qcor ) and quantile partial correlation ( qpcor ) . based on these two measures ,
we propose the quantile partial autocorrelation function ( qpacf ) and the quantile autocorrelation function ( qacf ) to identify the order of the qar model and to assess model adequacy , respectively .
it is noteworthy that the application of qcor and qpcor is not limited to qar models .
they can be used broadly as the classical correlation and partial correlation measures in various contexts .
the rest of this article is organized as follows .
section 2 introduces qcor and qpcor .
furthermore , the asymptotic properties of their sample estimators are established .
section 3 obtains qpacf and its large sample property for identifying the order of qar model .
in addition , the autoregressive parameter estimator and its asymptotic distribution are demonstrated .
moreover , qacf and its resulting test statistics , together with their asymptotic results , are provided to examine the model adequacy .
section 4 conducts simulation experiments to study the finite sample performance of the proposed methods , and also presents an empirical example to demonstrate usefulness .
finally , we conclude the article with a brief discussion in section 5 .
all technical proofs of lemmas and theorems are relegated to the appendix .
for random variables @xmath1 and @xmath2 , let @xmath3 be the @xmath4th unconditional quantile of @xmath2 and @xmath5 be the @xmath4th quantile of @xmath2 conditional on @xmath1 .
one can show that @xmath5 is independent of @xmath1 , i.e. @xmath6 with probability one , if and only if the random variables @xmath7 and @xmath1 are independent , where @xmath8 is the indicated function .
this fact has been used by @xcite and @xcite , and it also motivates us to define the quantile covariance given below . for @xmath9 , define @xmath10 where the function @xmath11 .
subsequently , the quantile correlation can be defined as follows , @xmath12 where @xmath13 . in the simple linear regression with the quadratic loss function
, there is a nice relationship between the slope and correlation .
hence , it is of interest to find a connection between the quantile slope and @xmath14 .
to this end , consider a simple quantile linear regression , @xmath15,\ ] ] in which one attempts to approximate @xmath5 by a linear function @xmath16 ( see @xcite ) , where @xmath17 $ ] .
then , we obtain the relationship between @xmath18 and @xmath19 given below .
[ lem1 ] suppose that random variables @xmath1 and @xmath2 have a joint density and @xmath20 .
then the values of @xmath21 are unique , and the quantity @xmath22 if and only if the quantile correlation @xmath23 .
it is noteworthy that the proposed quantile covariance here does not enjoy the symmetry property of the classical covariance , i.e. , @xmath24 .
this is because the first argument of the quantile covariance or the quantile correlation is related to the @xmath4th quantile , while the second argument is the same as that of the classical covariance .
accordingly , @xmath25 .
suppose that a quantile linear regression model has the response @xmath2 , a @xmath26 vector of covariate @xmath27 , and an additional covariate @xmath1 .
in the classical regression model , one can construct the partial correlation to measure the linear relationship between variables @xmath2 and @xmath1 after adjusting ( or controlling ) vector @xmath27 ( e.g. , see @xcite ) .
this motivates us to propose the quantile partial correlation function . to this end , let @xmath28 where @xmath29 is a vector of unknown parameters .
accordingly , @xmath30 is the linear effect of @xmath27 on @xmath1 .
next , consider @xmath31.\ ] ] as a result , @xmath32 is the linear effect of @xmath27 on the quantile @xmath2 ( i.e. , the linear approximation of @xmath33 ) .
it can also be shown that @xmath34 , @xmath35=0 $ ] and @xmath36=0 $ ] if the random vector @xmath37 satisfies the conditions stated in the forthcoming lemma [ lem2 ] . using these facts
, we define the quantile partial correlation as follows , @xmath38 } { \sqrt{(\tau-\tau^2)e(x-\alpha_1-\beta_1^{\prime}\mathbf{z})^2}}\\ & = \frac{e[\psi_{\tau}(y-\alpha_2-\beta_2^{\prime}\mathbf{z})x ] } { \sqrt{(\tau-\tau^2)\sigma_{x|\mathbf{z}}^2 } } , \end{split}\end{aligned}\ ] ] where @xmath39 .
this indicates that the covariate @xmath1 has no additional linear contribution to the quantile response @xmath2 if @xmath40 with probability one , where @xmath41.\ ] ] this leads to the following lemma .
[ lem2 ] suppose that the random vector @xmath37 has a joint density with @xmath20 and @xmath42 , where @xmath43 is the euclid norm .
then @xmath44 if and only if the quantile partial correlation @xmath45 .
since the true @xmath46 and @xmath47 are often unknown in practice , we introduce their sample versions given below .
suppose that the data @xmath48 are identically and independently generated from a distribution of @xmath49 .
let @xmath50 be the sample @xmath4th quantile of @xmath51 , where @xmath52 is the empirical distribution function .
based on equation ( [ eqc ] ) , the sample estimate of the quantile correlation @xmath19 is defined as @xmath53 where @xmath54 , and @xmath55 . to study the asymptotic property of @xmath56 , denote @xmath57 and @xmath58 as the density of @xmath2 and the conditional density of @xmath2 given @xmath1 , respectively .
in addition , let @xmath59 , @xmath60/f_y(q_{\tau , y})$ ] , @xmath61 , @xmath62 ^ 2-[\textrm{qcov}_{\tau}\{y , x\}]^2,\ ] ] @xmath63-\sigma_x^2\cdot\textrm{qcov}_{\tau}\{y , x\},\ ] ] and @xmath64,\ ] ] where @xmath65 is defined as in the previous subsection . then , we obtain the following result .
[ thm1 ] suppose that @xmath66 and there exists a @xmath67 such that the density @xmath57 is continuous and the conditional density @xmath58 is uniformly integrable on @xmath68 $ ] .
then @xmath69 to apply the above theorem , one needs to estimate the asymptotic variance @xmath70 . to this end
, we employ a nonparametric approach , such as the nadaraya - watson regression , to estimate the function @xmath71 , and denote it as @xmath72 .
we further assume that the random vector @xmath73 has a joint density , which leads to @xmath74 .
accordingly , we obtain the estimate , @xmath75 , where @xmath76 is the @xmath4th sample quantile of @xmath77 . finally , the rest of quantities , @xmath78 , @xmath65 , @xmath79 , @xmath80 , @xmath81 , and @xmath82 contained in @xmath70 can be , respectively , estimated by @xmath83 , @xmath84 , @xmath85 , @xmath86 , @xmath87 ^ 2 -[\widehat{\textrm{qcov}}_{\tau}\{y , x\}]^2 $ ] , and @xmath88 . as a result
, we obtain the estimate of @xmath70 , and denote it by @xmath89 .
we next estimate the quantile partial correlation @xmath90 .
let @xmath91 based on equation ( [ eqd ] ) , the sample quantile partial correlation is defined as @xmath92 where @xmath93 . to investigate the asymptotic property of @xmath94 ,
denote the conditional density of @xmath2 given @xmath27 and the conditional density of @xmath2 given @xmath27 and @xmath1 by @xmath95 and @xmath96 , respectively .
in addition , let @xmath97 , @xmath98 , @xmath99 , @xmath100 $ ] , @xmath101 $ ] , @xmath102 , @xmath103 , @xmath104 ^ 2-\{e[\psi_{\tau}(y-\theta_2^{\prime}\mathbf{z}^*)x]\}^2,\ ] ] @xmath105 -\sigma_{x|\mathbf{z}}^2\cdot e[\psi_{\tau}(y-\theta_2^{\prime}\mathbf{z}^*)x],\ ] ] and @xmath106)^2}{4\sigma_{x|\mathbf{z}}^6 } -\frac{\sigma_{25}\cdot e[\psi_{\tau}(y-\theta_2^{\prime}\mathbf{z}^*)x]}{\sigma_{x|\mathbf{z}}^4 } + \frac{\sigma_{24}}{\sigma_{x|\mathbf{z}}^2 } \right],\ ] ] where @xmath107 , @xmath108 , @xmath109 , @xmath110 and @xmath111 are defined as in the previous subsection .
then , we have the following result . [ thm2 ]
suppose that @xmath112 , @xmath113 , @xmath114 , @xmath115 , @xmath116 , and there exists a @xmath67 such that @xmath117 and @xmath118 are uniformly integrable on @xmath119 $ ]
. then @xmath120\rightarrow_dn(0,\omega_2).\ ] ] to estimate the asymptotic variance @xmath121 given in theorem 2 , we consider @xmath122 and @xmath123 $ ] .
in addition , assume that the random vector @xmath49 has a joint density .
we then have that @xmath124=f_{y^*}(0)\cdot e[x\mathbf{z}^*|y^*=0]$ ] , @xmath125 $ ] , and @xmath126\{e[\mathbf{z}^*\mathbf{z}^{*\prime}|y^*=0]\}^{-1}$ ] , where @xmath127 is the density of @xmath128 .
applying the same nonparametric technique as that used for estimating @xmath129 in theorem 1 , we could estimate each of the vector and matrix components in @xmath130 $ ] and @xmath131 $ ] , respectively , from the data @xmath132 , where @xmath133 .
accordingly , @xmath134^{-1}$ ] .
subsequently , the rest of quantities involved in @xmath121 , @xmath111 , @xmath135 , @xmath136 , @xmath137 , and @xmath138 can be , respectively , estimated by @xmath93 , @xmath139 , @xmath140 , @xmath141 ^ 2-[\widehat{\textrm{qcov}}_{\tau}\{y^*,x\}]^2 $ ] , and @xmath142 .
consequently , we obtain the estimate of @xmath121 , and denote it by @xmath143 .
it is noteworthy that the quantile correlation and quantile partial correlation can be broadly used as the classical correlation and partial correlation in regression analysis ( e.g. , variable selections ) , although our focus is on quantile autoregressive models .
suppose that @xmath144 is a strictly stationary and ergodic time series , and @xmath145 is the @xmath146-field generated by @xmath147 .
we then follow @xcite s ( 2006 ) approach and present qar models ; i.e. , conditional on @xmath148 , the @xmath4th quantile of @xmath149 has the form of @xmath150 where @xmath151s are unknown functions mapping from @xmath152\rightarrow r$ ] . following the box - jenkins classical approach
, we next introduce the qpacf of a time series to identify the order of a qar model , and then propose using the qacf of residuals to assess the adequacy of the fitted model . for the positive integer @xmath153 , let @xmath154 , @xmath155 , and @xmath156 $ ] , where the notations @xmath157 and @xmath158 are a slight abuse since they have been used to denote the regression parameters in section 2 . from equation ( [ eqd ] )
, we obtain the quantile partial correlation between @xmath149 and @xmath159 after adjusting the linear effect @xmath160 , @xmath161}{\sqrt{(\tau-\tau^2)e(y_{t - k}-\alpha_1-\beta_1^{\prime}\mathbf{z}_{t , k-1})^2}},\ ] ] and it is independent of the time index @xmath162 due to the strict stationarity of @xmath144 .
analogous to the definition of the classical pacf ( * ? ? ?
* chapter 2 ) , we name @xmath163 to be the qpacf of time series @xmath144 .
it is also noteworthy that @xmath164 .
we next show the cut - off property of qpacf .
[ lem3 ] if @xmath165 with @xmath166 , @xmath167 and @xmath168 ^ 2>0 $ ] , then @xmath169 , and @xmath170 for @xmath171 .
the above lemma indicates that the proposed qpacf plays the same role as that of pacf in the classical ar model identification . in practice
, one needs the sample estimate of qpacf . to this end , let @xmath172 and @xmath173 .
according to ( [ seqd ] ) , we obtain the estimation for @xmath163 , @xmath174 and we term it the sample qpacf of the time series . to study the asymptotic property of @xmath175 , we introduce the following assumption , which is similar to condition a.3 in @xcite .
[ assum3 ] @xmath167 , @xmath168 ^ 2>0 $ ] , and there exists a @xmath67 such that @xmath176 is uniformly integrable on @xmath119 $ ] .
furthermore , let @xmath177 by , the random variable @xmath178 is independent of @xmath159 for any @xmath179 , and @xmath180 for @xmath181 .
let @xmath176 be the conditional density of @xmath182 on the @xmath146-field @xmath148 , and @xmath183 .
moreover , let @xmath184 $ ] , @xmath185 $ ] , @xmath186 $ ] , @xmath187 $ ] , and @xmath188 then , we obtain the asymptotic result given below .
[ thm3 ] for @xmath171 , if @xmath189 , @xmath190 and assumption [ assum3 ] is satisfied , then @xmath191 and @xmath192 to estimate @xmath193 in the above theorem , we first apply the @xcite method to obtain the estimation of @xmath194 given below .
@xmath195 where @xmath196 is the estimated @xmath4th quantile of @xmath149 and @xmath197 is the bandwidth selected via appropriate methods ( e.g. , see @xcite ) .
afterwards , we can use the sample averaging to approximate @xmath198 , @xmath199 , @xmath200 , @xmath201 , @xmath202 , and @xmath203 by replacing their @xmath176 , @xmath107 , and @xmath108 , respectively , with @xmath204 , @xmath205 and @xmath206 .
accordingly , we obtain an estimate of @xmath193 , and denote it as @xmath207 . in sum , we are able to use the threshold values @xmath208 to check the significance of @xmath175 . to demonstrate how to use the above theorem to identify the order of a qar model , we generate the observations @xmath209 from @xmath210 where @xmath211 is the standard normal cumulative distribution function , @xmath212 , and @xmath213 is an @xmath214 sequence with uniform distribution on @xmath152 $ ] .
we attempt to fit the qar model with @xmath215 , 0.4 , 0.6 , and 0.8 , respectively , to the observed data @xmath144 .
figure [ fig1 ] presents the sample qpacf @xmath175 for each @xmath4 with the reference lines @xmath208 .
we may conclude that the order @xmath216 is 1 when @xmath215 and 0.4 , while @xmath216 is 0 when @xmath217 and 0.8 .
after the order @xmath216 of model is correctly identified , we subsequently fit the selected model to data .
let @xmath218 be an any parameter vector in model and @xmath219 be the true value of @xmath220 .
consider @xmath221 where @xmath222 .
in addition , let @xmath223 $ ] , @xmath224 $ ] , and @xmath225 .
we then obtain the following asymptotic property of the estimated parameter .
if @xmath226 and assumption [ assum3 ] is satisfied , then @xmath227 the above result is similar to that of theorem 2 in @xcite , although we make different assumptions .
the @xmath228 in the above theorem can be estimated by applying the same techniques used for the estimation of @xmath193 . for the errors @xmath229 defined in
, we employ equation ( [ eqc ] ) and the fact that @xmath230 , and obtain qacf between @xmath229 and @xmath231 as follows , @xmath232\ } } { \sqrt{(\tau-\tau^2 ) { \sigma}_e^2}},\ ] ] where @xmath233 .
suppose that the qar model is correctly specified .
we can show that @xmath234 for @xmath179 .
hence , we are able to use @xmath235 to assess the model fit . in the sample version ,
we consider the residuals of the qar model , @xmath236 for @xmath237 , and @xmath238 for @xmath239 .
it can be verified that the @xmath4th empirical quantile of @xmath240 is zero .
based on this fact and equation ( [ seqc ] ) , we obtain the estimation of @xmath235 , @xmath241 where @xmath153 is a positive integer , @xmath242 , @xmath243 , and the @xmath4th empirical quantile of @xmath240 is zero .
we name @xmath244 the sample qacf of residuals . adapting the classical linear time series approach @xcite
, we examine the significance of @xmath245 individually and jointly . for
the given positive integer @xmath246 , let @xmath247 , @xmath248 $ ] , @xmath249 $ ] , and @xmath250 then , we obtain the asymptotic distribution of @xmath251 given below .
[ thm5 ] assume that @xmath226 , @xmath252 , and assumption [ assum3 ] holds .
we then have @xmath253 applying the same techniques as used in the estimate of @xmath193 , we are able to estimate the asymptotic variance @xmath254 and denote it @xmath255 . in addition , let the @xmath153-th diagonal element of @xmath255 be @xmath256 .
then , one can employ @xmath257 to examine the significance of the @xmath153-th lag in the residual series . to check the significance of @xmath258 jointly ,
it is natural to consider the test statistic @xmath259 .
however , @xmath255 may not be invertible .
hence , we approximate @xmath254 by @xmath260 , which holds under the assumption that @xmath229 is an independent and identically distributed ( @xmath261 ) sequence and @xmath194 is a constant .
the resulting matrix is idempotent and has rank @xmath262 .
this allows us to obtain a box - pierce type @xcite test statistic , @xmath263 which follows an approximately chi - squared distribution with @xmath262 degrees of freedom , @xmath264 .
accordingly , @xmath265 can be used to test the significance of @xmath266 to @xmath267 jointly .
we conduct five simulation experiments to assess the finite - sample performance of the proposed methods . specifically , the first simulation experiment is for the sample quantile correlation and the sample quantile partial correlation proposed in section 2 , and the last four experiments are , respectively , for identification , estimation , and diagnosis as introduced in section 3 . in all experiments ,
we conduct 1000 realizations for each combination of sample sizes @xmath268 , 100 , and 200 and quantiles , @xmath269 , 0.50 , and 0.75 . in the first simulation experiment
, we generate the @xmath261 samples @xmath270 from the following multivariate normal distribution , @xmath271 after algebraic simplification , we obtain that @xmath272 ^ 2\}/\sqrt{(\tau-\tau^2)2\pi},\ ] ] and @xmath273 , where @xmath274 is the cumulative standard normal distribution .
tables [ table1 ] and [ table2 ] present the bias ( bias ) and estimated standard deviation ( esd ) , respectively , of the sample quantile correlations @xmath56 and the sample quantile partial correlations @xmath275 , calculated from 1000 realizations .
to estimate the asymptotic variances @xmath276 and @xmath277 , we mainly need to estimate the quantities @xmath278 and @xmath279 , addressed in subsection 2.2 . to this end
, we employ the nadaraya - watson approach with the two bandwidth selection methods proposed by @xcite and @xcite , respectively , which are given below .
@xmath280 ^ 2}\right\}^{1/5 } \hspace{5mm}\text{and}\hspace{5 mm } h_{hs}=n^{-1/3}z_{\alpha}^{2/3}\left\{\frac{1.5\phi^2(\phi^{-1}(\tau))}{2(\phi^{-1}(\tau))^2 + 1}\right\}^{1/3},\ ] ] where @xmath281 is the standard normal density function , @xmath282 , for the construction of @xmath283 confidence intervals , and @xmath284 is set to 0.05 .
furthermore , we consider two more bandwidths , @xmath285 and @xmath286 , suggested by @xcite . in sum
, we have four bandwidth choices .
the resulting asymptotic variance estimates , @xmath89 and @xmath143 , yield their correspondingly asymptotic standard deviations ( asds ) given in tables [ table1 ] and [ table2 ] .
both tables indicate that the asds are close to their corresponding esds even when @xmath268 , and they become smaller as the same size gets larger .
in addition , biases are close to zero , and decrease as the sample size increases . moreover ,
all four bandwidths lead to similar results , although @xmath286 is slightly better than the others .
the second monte carlo experiment studies the performance of qpacf for identifying the order of the qar models .
we generate the data from the following process , @xmath287 where @xmath288 is an @xmath214 sequence with standard normal distribution . under the above setting
, it can be shown that @xmath289 .
we then employ the approach of @xcite with the four bandwidths used in the first experiment to estimate the density function , @xmath194 .
this allows us to further estimate the variance matrix @xmath193 in theorem [ thm3 ] ( see subsection 3.1 ) .
table [ table3 ] presents the bias and estimated standard deviation of @xmath175 at @xmath290 , 4 , and 6 .
it shows that biases are small even when @xmath268 , and the esds are close to the asds as well as their theoretical value @xmath291 .
the third simulation experiment investigates the finite - sample performance of the qar estimates .
we use the same data generated from , and then fit it with the qar model with @xmath292 .
in addition , we employ the same approach as given in the second experiment to estimate @xmath194 . as a result
, the variance matrix @xmath228 in theorem [ thm4 ] can be estimated ( see subsection 3.1 ) .
table [ table4 ] presents the biases , estimated standard deviations , and asymptotic standard deviations of parameter estimates @xmath293 and @xmath294 .
it shows that biases are close to zero even when the sample size is as small as @xmath268 .
in addition , the esds are close to the asds , and both of them decrease as the sample size increases .
moreover , there is no discernible difference among the four bandwidths , although @xmath286 often yields the smallest asd .
the fourth simulation experiment examines the finite - sample performance of the sample qacf of residuals individually via the asymptotic result in theorem [ thm5 ] .
all settings are the same as those in the third experiment .
table [ table5 ] presents the biases , estimated standard deviations , and asymptotic standard deviations of @xmath244 at @xmath295 , 3 , and 5 .
apparently , biases are small and the asds are close to their corresponding esds . finally , the fifth experiment studies the approximate test statistic @xmath265 . to this end
, we generate data from the following process , @xmath296 where @xmath288 are @xmath261 standard normal random variables .
in addition , @xmath297 corresponds to the null hypothesis , while @xmath298 is associated with the alternative hypothesis .
moreover , the nominal level is 5% .
table [ table6 ] reports sizes and powers of @xmath265 with @xmath299 .
it shows that @xmath265 controls the size well , and its power increases quickly when the sample size or @xmath300 becomes larger .
consequently , the above six simulation studies perform satisfactorily and support our theoretical findings .
this example considers the log return ( as a percentage ) of the daily closing price on the nasdaq composite from january 1 , 2002 to december 31 , 2007 .
there are 1,235 observations in total , and figure [ nasdaq ] depicts the time series plot and the classical sample acf .
it is not surprising to conclude that these returns ( i.e. , log returns ) are uncorrelated and can be treated as an evidence in support of the fair market theory .
however , @xcite found that the stock markets under - react to good news in bad times and over - react to bad news in good times .
hence , @xcite proposed aligning a good ( bad ) state with upper ( lower ) quantiles by fitting their stock returns data with the qar(1 ) type models .
this motivates us to employ the general qar model with our proposed methods to explore the dependence pattern of stock returns at a lower quantile ( @xmath215 ) , the median ( @xmath301 ) , and an upper quantile ( @xmath302 ) .
we first fit the returns at the lower quantile ( @xmath215 ) , and then present its sample qpacf in panel a of figure [ nasdaq_fit ] .
it shows that lags 1 , 2 , and 13 are significant , which suggests qar(13 ) could be considered for model fitting .
we then refine the model via the backward variable selection procedure at the @xmath303 significance level .
the resulting model is @xmath304 where the subscripts of parameter estimates are their associated standard errors .
accordingly , the above coefficients are all significant at the 5% significance level . in addition , the second graph in panel a presents the sample qpacf of residuals , and no lags stand out .
this , together with the @xmath216-value of @xmath305 being 0.742 , implies that this model is adequate .
we next consider the scenario with @xmath301 .
the sample qpacf in panel b indicates that all lags are insignificant .
hence , we fit the following model , @xmath306 the above coefficient is not only small , but also not significant . in addition , none of the lags in the sample qacf of residuals in panel b show significance . moreover , the @xmath216-value of @xmath305 is 0.566 .
consequently , the above model is appropriate .
finally , we study the upper quantile scenario with @xmath302 .
the sample qpacf in panel c exhibits that lags 1 , 2 , 7 , 10 and 15 are significant , and suggests qar(15 ) could be considered for model fitting .
after refining the model via the backward variable selection procedure , we obtain @xmath307 where all coefficients are significant at the @xmath303 significance level .
in addition , the sample qacf of residuals in panel c displays that all lags are insignificant .
this , in conjunction with the @xmath216-value of @xmath305 being 0.215 , indicates that the above model fits the data reasonably well .
based on the three fitted qar models , ( [ naslow ] ) , ( [ nasmid ] ) , and ( [ nashigh ] ) , we obtain the following conclusions .
( i. ) the lag coefficients at the lower quantile ( @xmath215 ) are all positive .
this indicates that if the returns in past days have been positive ( negative ) , then today s negative return is alleviated ( even lower ) .
it also implies that stock markets under - react to good news in bad times .
the lag coefficients at the upper quantile ( @xmath302 ) are all negative .
this shows that if the returns in past days have been negative ( positive ) , then today s positive return ( @xmath302 ) is even higher ( dampened ) . as a result ,
stock markets over - react to bad news in good times .
the intercept at ( @xmath301 ) has a small value and is insignificant at the 5% significance level .
thus , the conditional median of returns is almost zero as we expected .
in addition , equation ( [ nasmid ] ) indicates that today s return is not affected by the returns of recent past days .
although we only report the results of the lower and higher quantiles at @xmath215 and @xmath302 , our studies yield the same conclusions across various lower and upper quantiles . in sum
, our proposed methods support veronesi s ( @xcite ) equilibrium explanation for stock market reactions .
in quantile regression models , we propose the quantile correlation and quantile partial correlation .
then , we apply them to quantile autoregressive models , which yields the quantile autocorrelation and quantile partial autocorrelation . in practice ,
the response time series may depend on exogenous variables .
hence , it is of interest to extend those correlation measures to the quantile autoregressive model with the exogenous variables given below .
@xmath308 where @xmath309 is a vector of time series , and @xmath310 and @xmath311 are functions @xmath152\rightarrow r$ ] , see @xcite . in addition , the application of the proposed correlations to the quantile regression model with autoregressive errors is worth further investigation .
clearly , the contribution of the proposed measures is not limited to those two models .
for example , variable screening and selection ( e.g. , @xcite ; @xcite ) in quantile regressions are other important topics for future research . in sum , this paper introduces valuable measures to broaden and facilitate the use of quantile models .
for @xmath312 , denote the function @xmath313 $ ] , where @xmath314 .
we first show that @xmath315 is a continuously differentiable function and has derivatives , @xmath316=p(y^*<0)-\tau \hspace{5mm}\text{and}\hspace{5 mm } \frac{\partial h(a , b)}{\partial b}=-e[\psi_\tau(y^*)x].\ ] ] for @xmath317 , @xmath318ds\\ & = -v\psi_{\tau}(u)+(u - v)[i(0>u >
v)-i(0<u < v ) ] , \end{split}\end{aligned}\ ] ] see @xcite .
this , together with hlder s inequality and the fact that @xmath319 is a continuous random variable , leads to @xmath320+e[\psi_\tau(y^*)x]|\\ & = |\frac{1}{c}e\{(y^*-cx)[i(0>y^*>cx)-i(0<y^*<cx)]\}|\\ & \leq e[|x|i(|y^*|<|c|\cdot|x|)]\leq ( ex^2)^{1/2}[p(|y^*|/|x|<|c|)]^{1/2},\end{aligned}\ ] ] which tends to zero as @xmath321 .
accordingly , @xmath322 is obtained .
analogously , we have @xmath323 . by hlder s inequality , we can further prove the continuity of @xmath324 .
moreover , the continuity of both @xmath1 and @xmath2 implies that @xmath325 is a continuous function .
it is noteworthy that @xmath315 is a convex function with @xmath326 .
this , in conjunction with the above results , demonstrates that the values of @xmath327 and @xmath18 satisfy @xmath328=0\hspace{5mm}\text{and}\hspace{5mm}e[\psi_\tau(y - a_0-b_0x)x]=0.\ ] ] we next show the uniqueness of @xmath21 .
suppose that there is another pair of values @xmath329 such that @xmath330 $ ] .
let @xmath331 and @xmath332 .
then , by and , @xmath333\\ & = -e[\xi \psi_{\tau}(y_0)]+e[(y_0-\xi)i(0>y_0>\xi)]+e[(\xi - y_0)i(0<y_0<\xi)]\\ & = e[(y_0-\xi)i(0>y_0>\xi)]+e[(\xi - y_0)i(0<y_0<\xi)].\end{aligned}\ ] ] note that both @xmath334 and @xmath335 are nonnegative random variables , and @xmath336 is a continuous random variable .
thus , with probability one , @xmath337 , which implies @xmath338 .
let @xmath343 and @xmath344.\ ] ] since the random vector @xmath345 has a joint density , we apply similar techniques to those in the proof of lemma [ lem1 ] to show that @xmath346=0 , \hspace{5 mm } e[\psi_{\tau}(y^*)\mathbf{z}]=\mathbf{0},\ ] ] and the values of @xmath347 , @xmath348 and @xmath349 are unique and satisfy @xmath350=\mathbf{0},\ ] ] where @xmath351 is a @xmath352 zero vector . from , if @xmath353 , then @xmath354=0 $ ] . on the other hand , @xmath355 , together with , implies that equation with @xmath353 holds .
accordingly , we have shown that @xmath355 if and only if @xmath353 .
based on the definitions of @xmath356 and @xmath357 in subsection 2.1 , we further have that @xmath358 , @xmath359 , and @xmath360 . finally , using the fact that @xmath361 completes the proof . for @xmath317
, we have that @xmath368 using this result , we then obtain @xmath369 where @xmath370 and @xmath371\\ & = i(\widehat{q}_{\tau , y}-{q}_{\tau , y}<y_i-{q}_{\tau , y}<0 ) -i(\widehat{q}_{\tau , y}-{q}_{\tau , y}>y_i-{q}_{\tau , y}>0).\end{aligned}\ ] ] we next consider the second term on the right - hand side of . for any @xmath374 , denote @xmath375\}x_i,\ ] ] where @xmath376= -\int_{{q}_{\tau , y}}^{{q}_{\tau , y}+n^{-1/2}v}f_{y_i|x_i}(y)dy\ ] ] and @xmath377 is the conditional density of @xmath378 given @xmath379 .
then , by hlder s inequality , we have that @xmath380 ^ 2&=e[g_{\tau}(y_i,{q}_{\tau , y},{q}_{\tau , y}+n^{-1/2}v)x_i]^2 \\ & \leq [ p(|y_i-{q}_{\tau , y}|<n^{-1/2}v)]^{1/2}[ex_i^4]^{1/2}=o(1 ) .
\end{split}\end{aligned}\ ] ] after algebraic simplification , we further obtain @xmath381\\ & = \frac{1}{\sqrt{n}}\sum_{i=1}^n|\{g_{\tau}(v_1^*)-g_{\tau}(v)\}x_i| + e[|\{g_{\tau}(v_1^*)-g_{\tau}(v)\}x_i||x_i],\end{aligned}\ ] ] where @xmath382 takes the value of @xmath383 or @xmath384 . hence , @xmath385 , \end{split}\end{aligned}\ ] ] where @xmath386 and @xmath387 when @xmath388 is large . both ( [ thm1_eq3 ] ) and ( [ thm1_eq4 ] ) , in conjunction with the theorem s assumptions and the finite converging theorem ,
imply that @xmath389 for any @xmath390 .
in addition , applying the theorem in section 2.5.1 of @xcite , we have @xmath391 accordingly , @xmath392}{f_y({q}_{\tau , y})}\cdot \frac{1}{\sqrt{n}}\sum_{i=1}^n\psi_{\tau}(y_i-{q}_{\tau , y})+o_p(1 ) . \end{split}\end{aligned}\ ] ] subsequently , using , , and , we obtain that @xmath393\\ & = \frac{1}{\sqrt{n}}\sum_{i=1}^n [ \psi_{\tau}(y_i-\widehat{q}_{\tau , y})(x_i-\bar{x})-\textrm{qcov}_{\tau}\{y , x\}]\\ & = \frac{1}{\sqrt{n}}\sum_{i=1}^n [ \psi_{\tau}(y_i-{q}_{\tau , y})(x_i-\mu_{x|y } ) -\textrm{qcov}_{\tau}\{y , x\}]+o_p(1 ) , \end{split}\end{aligned}\ ] ] where @xmath129 is defined in subsection 2.2 . since @xmath394 ^ 2 = o_p(n^{-1/2}),\ ] ] we further have that @xmath395+o_p(1).\ ] ] moreover , , , the central limit theorem , and the cramer - wold device , lead to @xmath396 where @xmath397 and @xmath80 , @xmath81 , and @xmath82 are defined in subsection 2.2 .
finally , following the delta method ( * ? ? ?
* chapter 3 ) , we complete the proof .
we first consider the term @xmath398 in @xmath94 .
let @xmath399 , @xmath400 , @xmath97 and @xmath401 , where @xmath157 and @xmath402 are defined in subsections 2.1 and 2.2 , respectively . by the assumptions of this theorem , we have that @xmath403^{-1}e(\mathbf{z}^*_ix_i)$ ] , @xmath404 , and @xmath405 according the law of large numbers
, we then have that @xmath406 we next consider the numerator in @xmath94 . for the sake of simplicity ,
let @xmath407 , @xmath98 , and @xmath408 , where @xmath409 is defined in the proof of lemma 2 , and @xmath158 and @xmath410 are defined in subsections 2.1 and 2.2 , respectively . under the theorem
s assumptions , we employ similar techniques to those used in the proof of lemma [ lem1 ] and given in @xcite to show that there exists a unique @xmath411 such that @xmath412=\mathbf{0}$ ] and @xmath413\}^{-1 } \cdot \frac{1}{\sqrt{n}}\sum_{i=1}^n\psi_{\tau}(y_i^*)\mathbf{z}^*_i+o_p(1).\ ] ] using a similar method to that for obtaining , we have that @xmath414 where @xmath415\\ & = i[(\widehat{\theta}_2-\theta_2)^{\prime}\mathbf{z}^*_i < y^*_i<0 ] -i[(\widehat{\theta}_2-\theta_2)^{\prime}\mathbf{z}^*_i > y^*_i>0].\end{aligned}\ ] ] for any @xmath416 , let @xmath417.\end{aligned}\ ] ] applying similar techniques to those for obtaining and , we can demonstrate that @xmath418 ^ 2 & = e[g_{\tau}(y_i,\mathbf{z}_i,\theta_2,\theta_2+n^{-1/2}\mathbf{v})x_i]^2 \\ & \leq \{p(|y_i^*|\leq n^{-1/2}|\mathbf{v}^{\prime}\mathbf{z}^*_i|)\}^{1/2}\cdot ( ex_i^4)^{1/2 } = o(1)\end{aligned}\ ] ] and , for any @xmath419 and @xmath420 , @xmath421.\end{aligned}\ ] ] this implies that @xmath422 for any @xmath390 .
note that , by , @xmath423 . as a result , @xmath424 where @xmath425 and @xmath426 $ ] are defined in subsection 2.2 .
this , together with , results in @xmath427 subsequently , by , , the central limit theorem , and the cramer - wold device , we obtain that @xmath428 \end{array}\right ) \rightarrow_d n(0,\sigma_2),\ ] ] where @xmath429 and @xmath136 , @xmath137 , and @xmath138 are defined as in subsection 2.2 .
finally , following the delta method ( * ? ? ?
* chapter 3 ) , we complete the proof .
we next study the numerator of @xmath175 .
let @xmath434 , and @xmath435 , where @xmath351 is the @xmath367 vector defined in the proof of lemma 3 , and @xmath436 and @xmath437 are defined in subsection 3.1 .
it is noteworthy that the series @xmath144 is fitted by model with order @xmath438 and the true parameter vector @xmath411 .
accordingly , @xmath439 and the parameter estimate of @xmath411 is @xmath440 .
then , using in the proof of theorem [ thm4 ] , we obtain that @xmath441\}^{-1 } \cdot\frac{1}{n}\sum_{t = k+1}^n\psi_{\tau}(e_{t,\tau } ) \mathbf{z}^*_{t , k-1}+o_p(n^{-1/2}).\ ] ] applying a similar approach to that used in obtaining , and then using the above result , we further have that @xmath442 y_{t - k}\\ & = -\frac{1}{n}\sum_{t = k+1}^n \int_0^{(\widetilde{\theta}_2-\theta_2)^{\prime}\mathbf{z}^*_{t , k-1 } } f_{t-1}(s)dsy_{t - k}+o_p(n^{-1/2})\\ & = -(\widetilde{\theta}_2-\theta_2)^{\prime } \cdot\frac{1}{n}\sum_{t = k+1}^n f_{t-1}(0)y_{t - k}\mathbf{z}^*_{t , k-1}+o_p(n^{-1/2})\\ & = -a_1^{\prime}\sigma_{31}^{-1 } \cdot\frac{1}{n}\sum_{t = k+1}^n\psi_{\tau}(e_{t,\tau } ) \mathbf{z}^*_{t , k-1}+o_p(n^{-1/2 } ) , \end{split}\end{aligned}\ ] ] where @xmath199 and @xmath201 are defined as in subsection 3.1 .
subsequently , using similar techniques to those for obtaining and the result from equation ( [ thm3_eq2 ] ) , we obtain that @xmath443 y_{t - k}\\ & = \frac{1}{n}\sum_{t = k+1}^n \psi_{\tau}(e_{t,\tau } ) [ y_{t - k}-a_1^{\prime}\sigma_{31}^{-1}\mathbf{z}^*_{t , k-1}]+o_p(n^{-1/2 } ) .
\end{split}\end{aligned}\ ] ] equations and ( [ thm3_eq4 ] ) , together with the central limit theorem for the martingale difference sequence , complete the proof of the asymptotic normality of @xmath175 . from lemma [ lem3 ] , we also have that @xmath191 . for any @xmath444 , denote @xmath445 where @xmath446 . applying and techniques similar to those in the proof of theorem 3.1 in @xcite
, we can show that @xmath447\mathbf{v}+o_p(1).\end{aligned}\ ] ] note that @xmath448 is a convex function with respect to @xmath449 .
by @xcite , we then have the bahadur representation as follows , @xmath450\}^{-1}\cdot \frac{1}{\sqrt{n}}\sum_{t = p+1}^n\psi_{\tau}(e_{t,\tau}^*)\mathbf{z}^*_{t , p}+o_p(1).\ ] ] this , in conjunction with the central limit theorem and the cramer - wold device , completes the proof . without loss of generality
, we assume that @xmath451 is observable .
then @xmath452 for @xmath453 .
we first consider the term @xmath454 in @xmath244 . by the ergodic theorem and the fact that @xmath455
, we can show that @xmath456 and @xmath457 where @xmath458 is defined in subsection 3.2 .
we next consider the numerator of @xmath244 .
using the fact that @xmath459 , we obtain @xmath460 + o_p(n^{-1})\\ & = \frac{1}{n}\sum_{t = k+1}^n \psi_{\tau}(y_t-\widetilde{{\mbox{\boldmath{$\phi$}}}}^{\prime}(\tau)\mathbf{z}^*_{t , p})e_{t - k,\tau } \\ & \hspace{5mm}-(\widetilde{{\mbox{\boldmath{$\phi$}}}}(\tau)-{{\mbox{\boldmath{$\phi$}}}}(\tau))^{\prime } \cdot\frac{1}{n}\sum_{t = k+1}^n \psi_{\tau}(y_t-\widetilde{{\mbox{\boldmath{$\phi$}}}}^{\prime}(\tau)\mathbf{z}^*_{t , p})\mathbf{z}^*_{t - k , p } + o_p(n^{-1/2 } ) .
\end{split}\end{aligned}\ ] ] applying similar techniques to those used in obtaining , we are able to show that @xmath461e_{t - k,\tau } \\ & = -\sigma_{51,k}\sigma_{41}^{-1}\cdot\frac{1}{n}\sum_{t = k+1}^n\psi_{\tau}(e_{t,\tau})\mathbf{z}^*_{t , p}+o_p(n^{-1/2}),\end{aligned}\ ] ] where @xmath462 is defined in subsection 3.1 and @xmath463 $ ] . in addition , using similar techniques to those in obtaining and the above result , we further obtain that @xmath464+o_p(n^{-1/2}).\end{aligned}\ ] ] analogously , we can verify that @xmath465 the above results , together with , ( [ thm5_eq2 ] ) , and the fact that @xmath455 , imply @xmath466+o_p(n^{-1/2}),\ ] ] and @xmath467+o_p(n^{-1/2}),\ ] ] where @xmath468 and @xmath469 are defined in subsection 3.2 .
subsequently , applying the central limit theorem for the martingale difference sequence and the cramer - wold device , we complete the proof .
the sample qpacf of daily closing prices on the nasdaq composite and the sample qacf of residuals from the fitted models for @xmath215 , 0.5 , and 0.8 .
the dashed lines in the left and right panels correspond to @xmath471 and @xmath472 , respectively.,title="fig : " ] the sample qpacf of daily closing prices on the nasdaq composite and the sample qacf of residuals from the fitted models for @xmath215 , 0.5 , and 0.8 .
the dashed lines in the left and right panels correspond to @xmath471 and @xmath472 , respectively.,title="fig : " ] the sample qpacf of daily closing prices on the nasdaq composite and the sample qacf of residuals from the fitted models for @xmath215 , 0.5 , and 0.8 .
the dashed lines in the left and right panels correspond to @xmath471 and @xmath472 , respectively.,title="fig : " ] | in this paper , we propose two important measures , quantile correlation ( qcor ) and quantile partial correlation ( qpcor ) .
we then apply them to quantile autoregressive ( qar ) models , and introduce two valuable quantities , the quantile autocorrelation function ( qacf ) and the quantile partial autocorrelation function ( qpacf ) .
this allows us to extend the classical box - jenkins approach to quantile autoregressive models .
specifically , the qpacf of an observed time series can be employed to identify the autoregressive order , while the qacf of residuals obtained from the fitted model can be used to assess the model adequacy .
we not only demonstrate the asymptotic properties of qcor , qpcor , qacf , and pqacf , but also show the large sample results of the qar estimates and the quantile version of the ljung - box test .
simulation studies indicate that the proposed methods perform well in finite samples , and an empirical example is presented to illustrate usefulness
. _ keywords and phrases : _ autocorrelation function ; box - jenkins method ; quantile correlation ; quantile partial correlation ; quantile autoregressive model |
the recent radial velocity detection of a planet in the habitable zone of the nearby m dwarf proxima centauri ( hereafter proxima b and proxima ) @xcite has spurred over a dozen theoretical papers speculating on the planet s atmosphere ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , habitability ( e.g. , * ? ? ?
* ; * ? ? ?
* ) , and orbital and formation histories ( e.g. , * ? ? ?
* ; * ? ? ?
* ) as well as prospects for a direct detection or atmospheric characterization ( e.g. , * ? ? ?
* ; * ? ? ?
as proxima is the nearest neighbor to the solar system , it has been suggested as a target for future space missions , including those hoping to characterize its atmosphere and search for life ( e.g. , * ? ? ?
* ; * ? ? ?
in many of these studies , authors have assumed a rocky planet with a thin atmosphere or no atmosphere at all , and some have assumed a mass near or equal to the projected mass of @xmath7 @xmath1 , but little has been done to assign a degree of certainty to these assumptions .
most notably , previous studies have revealed two distinct populations of exoplanets with super - earth radii : ` rocky ' planets composed almost entirely of rock , iron , and silicates with at most a thin atmosphere , and ` sub - neptune ' planets which must contain a significant amount of ice or a h / he envelope ( e.g. , * ? ? ?
* ; * ? ? ?
if there is a significant probability that proxima b is of the latter composition , then this should be taken into account when assessing its potential habitability or observability . in this letter
, we generate posterior distributions for the mass of proxima b using monte carlo simulations of exoplanets with an isotropic distribution of inclinations , where the radii , masses , and compositions of the simulated planets are constrained by results from combined transit and radial velocity measurements of previously detected exoplanets . by comparing the posterior mass distribution to the composition of planets as a function of mass
, we determine the likelihood that proxima b is , in fact , a rocky world with a thin ( if any ) atmosphere .
radial velocity and transit studies of exoplanets have yielded mass and radius measurements for a statistically significant number of targets , thereby enabling the study of how the occurrence and composition of exoplanets varies with planet radii , orbital periods , and host star type . in this section ,
we review previous results which we will use to place stronger constraints on the mass and composition of proxima b. it can be shown ( e.g. , * ? ? ?
* ) that the probability distribution of @xmath8 corresponding to an isotropic inclination distribution is @xmath9 since this distribution peaks at @xmath10 , the mass distribution of an rv - detected planet - assuming no prior constraints on the mass - peaks at the minimum mass @xmath11 . in their models of the possible orbital histories of proxima b ,
@xcite find that galactic tides could have inflated the eccentricity of the host star s ( at the time unconfirmed ) orbit around the @xmath12 cen binary , leading to encounters within a few hundred au and the possible disruption of proxima s planetary system .
if so , this could affect the likely inclination of the planet in a non - isotropic way .
however , @xcite have presented radial velocity measurements showing that proxima is gravitationally bound to the @xmath12 cen system with an orbital period of 550,000 years , an eccentricity of @xmath13 , and a periapsis distance of 4,200 au . at this distance ,
the ratio of proxima s gravitational field to that of @xmath12 cen at the planet s orbit ( @xmath14 au ) is greater than @xmath15 ; unless proxima s orbit was significantly more eccentric in the past , it seems unlikely that @xmath12 cen would have disrupted the system .
@xcite provide up - to - date occurrence rates of planets around m dwarf stars from the _ kepler _ mission .
the sample is limited to @xmath16 days , over which they find the occurrence rates to be mostly independent of the period .
the binned rates and a regression curve , as well as their uncertainties , are presented in figure [ fig : occurrence_rates ] .
_ kepler _ statistics for m dwarfs remain incomplete below 1 @xmath6 , but complete statistics for earlier - type stars suggest a flat distribution for @xmath17 @xmath6 @xcite . since mass - radius relationships typically find a strong dependence of mass on radius ( @xmath18 ) ( e.g. * ? ? ?
* ; * ? ? ?
* ) , we assume _ a priori _ that proxima b ( @xmath19 @xmath1 ) is larger than @xmath20 @xmath6 . therefore , for this letter we adopt the regression curve fitted to the binned data , but set the occurrence rates to be flat for @xmath21 @xmath6 . * occurrence rates for m dwarf planets + ( with @xmath22 days ) * ( dotted ) to be flat , since the sample is incomplete in this range .
bottom : mass - radius relationships for the rocky ( blue ) and sub - neptune ( red ) populations .
the plotted relationships are from @xcite ( solid ) and @xcite ( dashed).[fig : occurrence_rates ] [ fig : mass - radius_relationship ] ] * mass - radius relationships for @xmath23 @xmath24 * ( dotted ) to be flat , since the sample is incomplete in this range .
bottom : mass - radius relationships for the rocky ( blue ) and sub - neptune ( red ) populations .
the plotted relationships are from @xcite ( solid ) and @xcite ( dashed).[fig : occurrence_rates ] [ fig : mass - radius_relationship ] ] multiple works ( e.g. * ? ? ?
* ; * ? ? ?
* ) have determined the existence of two distinct populations of exoplanets smaller than neptune ( @xmath25 @xmath6 ) : a small radius population with densities consistent with an entirely iron and silicate composition ( hereafter ` rocky ' ) , and a large radius population with lower density planets which must have significant amounts of ice or a thick h / he atmosphere ( hereafter ` sub - neptunes ' ) .
@xcite studies the abundance of planets of each composition as a function of radius .
they define @xmath26 as the likelihood that a planet of radius @xmath27 is dense enough to be consistent with a rocky composition , and determine @xmath28 for a sample of planets with known masses and radii . they suggest fitting the data with a two - parameter linear model : @xmath29 they find a step function to best describe the data , with @xmath30 fixed at zero and @xmath31 @xmath6 . for the purposes of this letter
, we prefer this fit , but will also vary @xmath32 and @xmath30 to see how they affect our results .
we stress that a planet for which @xmath33 is only _ sufficiently _ dense to be rocky ; we still can not necessarily exclude an ice or volatile component . here
, we will assume that all planets for which @xmath33 follow the low - radius m - r relationships given in the following section , which were empirically fitted without prior knowledge of the planets compositions .
for simplicity , we refer to these as ` rocky ' planets , and the other population as ` sub - neptunes ' , but we will revisit this distinction later on . since proxima b is in the habitable zone , it receives an amount of stellar flux comparable to that received by earth , so we should bear in mind the possibility that the volatile envelope of a sub - neptune could be lost due to photoevaporation .
@xcite model rocky planets with thick h / he envelopes in the habitable zones of m dwarfs , finding that planets with @xmath34 @xmath1 maintain their envelopes over gyr timescales and are therefore uninhabitable .
the 2@xmath35 lower limit on the minimum mass of proxima b is @xmath36 @xmath1 , so it is unlikely that any h / he envelope on the planet would evaporate under this rule .
however , we note that this study focuses on planets with a primarily rocky composition , so it may not be directly applicable to habitable zone sub - neptunes . additionally , @xcite empirically define boundaries for atmospheric evaporation as a function of stellar heating , escape velocity , and atmospheric composition . in particular , a planet receiving an earth - like flux must have an escape velocity above @xmath37 km / s in order to maintain an h@xmath38 atmosphere for 5 gyr .
we will revisit this requirement in section [ sec : escape ] .
empirically derived relationships between exoplanet masses and radii rely on radial velocity ( rv ) or transit - timing variation ( ttv ) measurements of transiting exoplanet masses .
@xcite fit a mass - radius ( hereafter m - r ) relationship to a sample of 65 transiting exoplanets , in which they find evidence for the two populations discussed in section [ sec : compositions ] .
through least - squares regression , they find the densities of the rocky planets to increase linearly with planet radius : @xmath39 while the rv - measured masses of sub - neptunes increase nearly linearly with planet radius : @xmath40 @xcite use an expanded version of this data set to fit power law m - r relationships using a more statistically robust bayesian method .
for the rocky planets , they find @xmath41 and for the sub - neptunes with rv - measured masses , @xmath42 due to the larger sample size and more robust fitting procedure , we adopt equations [ equation : wolfgang_rocky ] and [ equation : wolfgang_rv ] as our preferred m - r relationships , but for completeness we consider the @xcite relationships as well . we find that the choice of m - r relationships has a minimal impact on our final results .
both sets of relationships are plotted in figure [ fig : mass - radius_relationship ] .
it is important to note that the above relationships for sub - neptunes exclude masses measured by ttv , since ttv masses have been found to be systematically lower than rv masses .
this could indicate a selection bias or systematic error in the method used , but since proxima b was detected through rv measurements , we believe it is proper to exclude the ttv masses either way .
it is also clear that there is a significant spread in the masses of the observed planets .
@xcite suggest a spread of @xmath43 @xmath1 for the sub - neptune planets , which we adopt for our simulations . for rocky planets ,
the spread is noticeably smaller .
there are too few planets to constrain this spread , but it should most likely increase with mass , so we arbitrarily define the spread to be 30% of the calculated mass .
the fitted occurrence rates and their uncertainties ( @xmath44 ) are given in even bins in log - space .
we use them to generate a random sample of radii , where the number of radii in each bin ( @xmath45 ) is selected from a normal distribution with mean value @xmath46 and standard deviation @xmath47 .
we find that the results converge for 1,000 samples of the occurrence rates , with each sample containing @xmath48 radii . to each radius ,
we assign a composition ( ` rocky ' or ` sub - neptune ' ) based on the model of @xcite ( equation [ equation : rogers_linear ] ) , with @xmath49 @xmath6 and @xmath50 .
we then assign a mass to each radius and composition from a gaussian distribution with mean value @xmath51 - calculated using our chosen m - r relationships ( equations [ equation : wolfgang_rocky ] and [ equation : wolfgang_rv ] ) - and a standard deviation @xmath52 which represents the spread .
we choose a spread proportional to the calculated mass for rocky planets ( @xmath53 ) , but a constant spread for sub - neptunes ( @xmath54 @xmath1 ) .
we also reject negative masses , which could in principle bias the assigned masses towards higher - than - average values - however , we find that only a negligible number of masses are rejected . finally , we assign a line - of - sight inclination parameter @xmath8 to each planet , drawn from the isotropic probability distribution discussed in section [ sini_distribution ] . the prior mass and radius distributions , @xmath55 and @xmath56 , can be derived directly from the simulated sample .
factoring in the projected minimum mass @xmath11 , the posterior distributions @xmath57 and @xmath58 can be calculated from bayes formula : @xmath59 where @xmath60 represents mass or radius .
since @xmath11 is known , @xmath61 is just a normalizing constant .
taking @xmath62 @xmath1 as the projected mass of proxima b and the upper limit @xmath63
@xmath1 as its standard deviation , we calculate for each simulated planet @xmath64 then @xmath65 and @xmath66 are the average values of @xmath67 for each bin in mass or radius . the prior and posterior distributions are calculated for each sample of @xmath68 planets , and the final results are taken to be the mean result of 1,000 samples . the prior probability that a planet in a given mass bin is rocky is equal to the number of simulated rocky planets in that bin divided by the total number of planets in the same bin . since we want to know the likelihood that proxima b belongs to the ` rocky ' population , we multiply this prior composition probability distribution by the posterior mass distribution from the previous section and integrate over all masses .
* prior mass distribution * . [ fig : prior_posterior_pdf ] ] * posterior mass distribution * .
[ fig : prior_posterior_pdf ] ] * cumulative mass probability distribution * ( dashdot ) . the dotted lines intersect
68% and 95% confidence upper limits on the mass .
[ fig : cumulative_pdf ] ] the prior and posterior mass probability distributions for proxima b are plotted in figure [ fig : prior_posterior_pdf ] .
the shaded regions demonstrate the relative contributions of the populations at each mass .
the prior distribution is valid for rv - detected planets around m dwarfs with intermediate periods ( @xmath69 days ) and radii ( @xmath70 @xmath6 ) , while the posterior distribution can be taken as the mass probability distribution for proxima b. for reference , we include the posterior distribution given _ no _ prior constraints on the mass ; that is , the distribution resulting from an isotropic @xmath8 distribution and the measured @xmath11 with its uncertainty .
we find that this nearly matches our result , since both @xmath71 and @xmath55 are bottom - heavy .
figure [ fig : cumulative_pdf ] shows the cumulative probability that @xmath72 for both of the considered m - r relationships ( section [ sec : mass - radius_relationships ] ) as well as for the case of no prior mass distribution .
we find that there is little difference between the results for each m - r relationship . in order to verify that sub - neptune planets can maintain h@xmath38 envelopes in the habitable zone
, we compare the escape velocities of our simulated sub - neptunes to the @xmath37 km / s cutoff for h@xmath38 atmospheric escape ( assuming an earth - like stellar flux ) defined by @xcite . in both the prior and posterior distributions of escape velocities
, we find that fewer than 1% of the sub - neptunes have escape velocities below this threshold , with most having @xmath73 km / s .
therefore , we do not believe that proxima b will be subject to significant atmospheric loss if it has a sub - neptune composition .
table [ tab : cases ] lists the sets of parameters for which we run the simulation , including the mass spread @xmath52 for each composition and the central value ( @xmath32 ) and width ( @xmath30 , if nonzero ) of the transition region defined by equation [ equation : rogers_linear ] .
the following results for each case are given : the probability @xmath74 that proxima b belongs to the ` rocky ' category of planets , i.e. that its density is _ consistent _ with a fully iron and silicate composition , and the expectation values @xmath75 and @xmath76 of the mass and radius under the assumption that it belongs to this population .
case a is most consistent with the previous work we have cited , so we take it as our primary result . in this case
, there is a @xmath2 probability that proxima b belongs to the ` rocky ' population , with an @xmath77 likelihood that it belongs to the ` sub - neptune ' population . in the case
that it is rocky , the expectation values ( and 95% confidence intervals ) for the mass and radius are @xmath4 @xmath1 and @xmath5 @xmath6 .
we investigate the effect of increasing ( case b ) and decreasing ( case c ) the mass spread for each composition , which results in lower and higher values of @xmath74 , respectively .
this results from low - radius ( @xmath78 @xmath6 ) , low - mass sub - neptunes ; when @xmath52 is large , they can lie significantly below the m - r relation with masses between 1 and 2 @xmath1 , so that they are indistinguishable from the rocky planets in the mass domain . in cases d and e , we determine the effect of raising or lowering the threshold radius @xmath32 at which the rocky and sub - neptune populations are split .
a 0.2 @xmath6 offset in either direction , which encompasses most of the values suggested in the literature , results in a @xmath79 to @xmath80 shift in @xmath74 , where higher threshold radii allow for more rocky planets and therefore a higher probability of a rocky composition . furthermore , allowing for a non - zero width @xmath30 to the cutoff region allows sub - neptunes to exist with lower radii and masses , thereby decreasing @xmath74 . in all cases ,
we find @xmath74 to be between @xmath81 and @xmath82 using the @xcite m - r relationship , and we find similar values using the @xcite relationship ( e.g. @xmath83 for case a ) , so this result does not vary substantially over the range of reasonable values for the input parameters .
by considering occurrence rates from the _ kepler _ mission and empirically derived m - r relationships , we derive a posterior probability distribution for the actual mass of proxima b. if the planet has a rocky composition , i.e. if it obeys the low - radius m - r relationship of @xcite , then the expectation values of the mass and radius ( with 95% confidence intervals ) are @xmath4 @xmath1 and @xmath5 @xmath6 . in all of our simulations , we find a probability of 80% to 95% that proxima b belongs to the ` rocky ' population of planets defined in section [ sec : compositions ] . in our ` best guess ' scenario ( case a ) , this probability is 90% .
critically , we note that we have assumed all planets with @xmath33 ( according to the @xcite criterion ) are rocky planets , while in reality their density is only consistent with such a composition . with this in mind
, it is safest to say that there is _ at least _ a 10% chance that proxima b has a sub - neptune composition .
if it is a sub - neptune , then its surface gravity is high enough that it could maintain a thick hydrogen atmosphere . for future theoretical work involving the habitability and detectability of proxima b , we advise caution regarding assumptions made about its mass or composition
; if proxima b does possess a thick h / he envelope , then it is likely not habitable in the traditional sense .
even if the mass could be further constrained , sub - neptunes have been measured with masses as low as @xmath84 @xmath1 , so the composition can not be conclusively inferred from the mass alone .
nevertheless , the rocky composition originally asserted by @xcite remains the most likely possibility .
the results reported herein benefited from collaborations and/or information exchange within nasa s nexus for exoplanet system science ( nexss ) research coordination network sponsored by nasa s science mission directorate .
we thank benjamin rackham and gijs mulders for their constructive advice and insights , and the anonymous referee for their comments .
cccccccccc * case a * & & @xmath85 & * 1.9 @xmath86 * & * 1.5 @xmath24 * & * - * & & @xmath87 & @xmath88 @xmath86 & @xmath89 @xmath24 + case b & & @xmath90 & 3.8 @xmath1 & 1.5 @xmath6 & - & & @xmath91 & @xmath92 @xmath1 & @xmath93 @xmath6 + case c & & @xmath94 & 0.7 @xmath1 & 1.5 @xmath6 & - & & @xmath95 & @xmath96 @xmath1 & @xmath97 @xmath6 + case d & & @xmath98 & 1.9 @xmath1 & 1.7 @xmath6 & - & & @xmath99 & @xmath100 @xmath1 & @xmath101 @xmath6 + case e & & @xmath98 & 1.9 @xmath1 & 1.3 @xmath6 & - & & @xmath102 & @xmath103 @xmath1 & @xmath104 @xmath6 + case f & & @xmath98 & 1.9 @xmath1 & 1.5 @xmath6 & 1.2 @xmath6 & & @xmath105 & @xmath106 @xmath1 & @xmath107 @xmath6 + case g & & @xmath90 & 3.8 @xmath1 & 1.5 @xmath6 & 1.2 @xmath6 & & @xmath108 & @xmath109 @xmath1 & @xmath110 @xmath6 + | recent studies regarding the habitability , observability , and possible orbital evolution of the indirectly detected exoplanet proxima b have mostly assumed a planet with @xmath0 @xmath1 , a rocky composition , and an earth - like atmosphere or none at all . in order to assess these assumptions , we use previous studies of the radii , masses , and compositions of super - earth exoplanets to probabilistically constrain the mass and radius of proxima b , assuming an isotropic inclination probability distribution .
we find it is @xmath2 likely that the planet s density is consistent with a rocky composition ; conversely , it is at least @xmath3 likely that the planet has a significant amount of ice or an h / he envelope .
if the planet does have a rocky composition , then we find expectation values and 95% confidence intervals of @xmath4 @xmath1 for its mass and @xmath5 @xmath6 for its radius . |
it has long been clearly understood that the statement of the second law of thermodynamics concerning the increase in entropy in an isolated system as it goes to equilibrium refers only to the average behavior , but this was sufficient as long as one dealt only with macroscopic systems characterized by extremely narrow ensemble distributions with fluctuations that were essentially never observed .
more recently , with the ability to quantitatively monitor systems on the extremely small scales of single molecules and quantum dots , it is possible to study fluctuations around the average behavior . _ fluctuation theorems _ that hold arbitrarily far from equilibrium have thus become subject to experimental verification @xcite .
these theorems in general deal with the ratio of the probabilities of a given system trajectory and that of its time reversed trajectory , either as the system goes to equilibrium or as it evolves to a steady state under the action of nonequilibrium constraints imposed on the system . from this one can calculate , for example , the relative probabilities that the entropy of an isolated system away from thermodynamic equilibrium will spontaneously increase or decrease over a given period of time .
the ratio is essentially infinite in a macroscopic system away from equilibrium and is unity due to fluctuations in equilibrium , but in sufficiently small systems away from equilibrium it is merely large ( and experimentally accessible ) rather than infinite .
fluctuation theorems can take different forms depending on the specific problem under consideration , but they are all ultimately connected to the probabilistic asymmetry of system trajectories and time reversed trajectories .
equilibrium corresponds to the situation where the symmetry is restored as stated by the principle of microreversibility .
fluctuation theorems have been formulated for a wide range of dynamics such as driven hamiltonian dynamics @xcite , stochastic dynamics @xcite , deterministic thermostated dynamics @xcite , and even quantum dynamics @xcite . here
we focus on stochastic dynamics , in an effort to explore the validity of fluctuation theorems beyond the stochastic dynamics that have been considered to date . in this narrower context of stochastic dynamics
, most previous studies of fluctuation theorems have focused on systems described by markovian master equations or fokker - plank equations .
recently there have been some efforts to investigate fluctuation theorems for systems described by nonlinear generalized langevin equations @xcite with an external driving force as a nonequilibrium constraint .
our focus is on nonequilibrium systems described by continuous time random walks ( ctrw ) @xcite in which transitions between microscopic states may be caused by more than one mechanism .
the nonequilibirum constraint is imposed when these mechanisms have different statistical properties such as , for example , through contact with two heat baths maintained at different temperatures . in general , identifying such nonequilibrium constraints may itself be complicated @xcite , and we will here explicitly point to these differences .
we pose the following question : what properties of a ctrw are necessary for an entropy fluctuation theorem to be valid under such nonequilibrium constraints ? we note , for example , that ctrws are known to display aging @xcite as well as nonergodic phenomena @xcite which may significantly alter the behavior of the system under time reversal and prevent a fluctuation theorem from being satisfied . at the same time , ctrws under certain conditions reduce to markovian master equations which are known to satisfy fluctuation theorems .
ctrws therefore provide a good framework to study the validity of fluctuation theorems .
in particular , our results will hopefully contribute clarification to recent observations of anomalous statistics in the nonequilibrium fluctuations of single molecules and quantum dots @xcite .
a second purpose of this paper is the formulation of a general framework for the calculation of ( experimentally accessible ) counting statistics of events associated with a given mechanism .
examples of such events might involve particle or energy transfer . to accomplish this we use a method based on the propagation of the generating function associated with the probability distribution of the events , in the spirit of the method used for markovian master equations @xcite .
this will allow us to investigate the long - time behavior of the moments of the distribution associated with the counting statistics .
our basic ctrw model is constructed as follows .
we consider a stochastic dynamics between microscopic states @xmath9 of a system with a finite number of states .
the transitions between states may be due to different mechanisms @xmath10 .
for example , we will subsequently consider a system in which each microscopic state @xmath9 is characterized by a number @xmath11 of particles and an energy @xmath12 , and where the transitions between the @xmath9 s are triggered by different reservoirs ( heat baths ) @xmath10 .
suppose that the system arrives at state @xmath13 at a given time and that its next jump is to state @xmath9 at a time @xmath14 later via mechanism @xmath10 .
the distribution of waiting times ( wtd ) for this to occur is denoted by @xmath15 , with other related quantities specified in more detail subsequently .
we focus on waiting time distributions whose long - time behavior is reflected in the small-@xmath16 laplace transform @xmath17 where @xmath18 and @xmath19 .
the @xmath20 are elements of an arbitrary matrix .
a detailed discussion surrounding this choice can be found in @xcite .
when @xmath6 the long - time decay of the wtds is then of the power law form @xmath21 .
when @xmath1 the decay is at least as fast as @xmath22 but may be faster . in sec .
[ ctrw ] we present the ctrw model for an open system driven by different mechanisms described by different statistical properties , and formally express the probability that the system is in state @xmath9 at time @xmath14 . in sec .
[ gme ] we derive the generalized master equation satisfied by this probability and study its long time behavior . in sec .
[ gfcurrentsec ] we present a generating function formalism to calculate the probability distribution of heat and matter transfers which is used to study the long time behavior of the moments of the distribution . the principal results of this work , namely , the conditions for the validity of fluctuation theorems , are presented in sec . [ ptraj ] .
in particular , we show that fluctuation theorems for the entropy change can only be obtained if @xmath1 , that is , if the wtds decay at least as fast as @xmath2 at long times .
furthermore , even in this case the entropy change can be expressed as a familiar ratio of the probability of a trajectory and its time reversed trajectory only if the wtds satisfy constraints of separability .
a summary of results and some concluding remarks are presented in sec .
our goal in this section is to construct the probability that the system will be found in a particular state at time @xmath14 .
suppose that the system arrives at state @xmath13 at time zero , and that its next jump is to state @xmath9 at time @xmath14 via mechanism @xmath10 .
the waiting time distribution ( wtd ) @xmath15 for this event introduced earlier satisfies the normalization condition @xmath23 for convenience , we define @xmath24 .
the probability that no transition occurs up to time @xmath14 after arrival at @xmath13 at time zero is @xmath25 we see that by construction @xmath26 , that is , a jump eventually occurs with certainty .
we define the auxiliary distributions @xmath27 @xmath28 is the waiting time distribution of the first jump from state @xmath13 to any other state via a mechanism @xmath10 , and @xmath29 is the waiting time distribution of the first jump from state @xmath13 to any other state regardless of the mechanism , given that arrival at @xmath13 occurred at time zero .
we also define @xmath30 here @xmath31 is the probability that , being at @xmath13 , the next jump will be from state @xmath13 to @xmath9 via mechanism @xmath10 , and @xmath32 is the probability that the jump will be from @xmath13 to @xmath9 irrespective of the mechanism . note that by definition @xmath33 .
we finally define the probability @xmath34 that the next jump from @xmath13 will be due to mechanism @xmath10 , @xmath35 the normalization condition ( [ normwtd ] ) implies @xmath36 the determining feature of the problem for our purposes is the behavior of the first moment of the waiting time distribution @xmath15 and , in particular , whether it is finite or infinite .
this first moment is just the average time that the system remains in state @xmath13 before jumping to @xmath9 via mechanism @xmath10 : @xmath37 associated mean waiting times are given by @xmath38 the first is the average time that the system remains in state @xmath13 without jumping anywhere else via mechanism @xmath10 .
the second is the average time that the system remains at @xmath13 without making any jumps at all by any mechanism .
if @xmath39 , all moments of the waiting time distribution ( including the first moments or mean waiting times ) are divergent , whereas for @xmath40 the first moments @xmath41 are finite .
in this case , from ( [ s0wtdpsigeneral ] ) and ( [ tmoyenentre2 m ] ) it follows that @xmath42 .
suppose that we begin our observations at time @xmath43 , at which time we find the system in state @xmath13 .
since in general we may not know when the jump occurred that brought the system to that state , we need to distinguish the waiting time distribution of the first jump after time zero from that of subsequent jumps .
we mark this first waiting time distribution with a prime , @xmath44 .
the probability that no transition away from @xmath13 occurs up to time @xmath14 must be similarly distinguished , @xmath45 .
the primed functions are equal to the unprimed ones only if a jump occurred exactly at time zero or if the wtds decay exponentially . if there is no information about when the last jump before time @xmath43 occurred , then the primed functions can be related to the unprimed ones _ only _ if the mean waiting times @xmath41 are finite . in this case , an average over the uncertain past yields @xmath46 note that summing eq .
( [ wtfirstjump ] ) over @xmath9 and @xmath10 and using eq .
( [ nojumpwt ] ) leads to @xmath47 similarly , @xmath48 it is straightforward to construct an integral equation for @xmath49 , the probability that the system is in state @xmath9 at time @xmath14 . for this purpose
we also define @xmath50 , the probability that the system jumps onto state @xmath9 at time @xmath14 .
the following integral ctrw relations are evident @xcite : @xmath51 and @xmath52 since these are convolutions , it is easiest to solve for the laplace transform @xmath53 of the desired probability @xmath49 .
the laplace transforms of these two relations are @xmath54 and @xmath55 the solution is most neatly expressed in terms of the matrices @xmath56 and @xmath57 with matrix elements @xmath58_{mm ' } \equiv \tilde{\psi}_{m , m'}'(s ) ,
\quad [ \tilde{\phi}(s)]_{mm ' } \equiv \delta_{mm ' } \tilde{\phi}_{m}(s ) , \label{ctrwformalsol1}\ ] ] and the vectors @xmath59 and @xmath60 with elements @xmath61 and @xmath53 .
the solution of eq .
( [ ctrw1lap ] ) is @xmath62 where @xmath63 is the identity matrix . using ( [ ctrw2lap ] )
, the formal solution of the ctrw then is @xmath64 if @xmath65 , we get the simplified form @xmath66
in this section we construct the generalized master equation for the probability that the system is in state @xmath9 at time @xmath14 .
this equation facilitates the exploration of the behavior of this probability at long times . using the laplace transform of ( [ nojumpwt ] ) , @xmath67 we can rewrite ( [ ctrw2lap ] ) as @xmath68 inserting ( [ ctrw1lap ] ) in the second term of ( [ ctrw3lap ] ) and using eq .
( [ ctrw2lap ] ) , we obtain @xmath69 where the transition matrix elements are given by @xmath70 and @xmath71 the inhomogeneous term @xmath72 in eq .
( [ generalizedmelaplace ] ) thus depends on the initial condition .
it vanishes if a jump occurs at time zero or if the waiting time distributions are exponential because then @xmath73 . upon inverse laplace transformation
we arrive at the generalized master equation @xmath74 the generalized master equation is clearly non - markovian unless @xmath75 is independent of @xmath16 .
this occurs , for example , for separable distributions @xmath76 with @xmath77 . indeed , in this case @xmath78 . of interest for our purposes
is the long - time behavior of the probability @xmath49 .
here we distinguish the case @xmath1 , associated with finite mean waiting times , from the case @xmath6 , associated with divergent mean waiting times .
consider first the long time behavior of the generalized master equation in the case @xmath1 . from eq .
( [ laplacephi ] ) it follows that @xmath79 where @xmath80 . since it follows from this result and from eq .
( [ s0wtdpsigeneral ] ) that @xmath81 and @xmath82 , eq .
( [ rate ] ) then immediately leads to @xmath83 since @xmath84 is constant when @xmath1 , we find that @xmath85 is also constant [ see eq .
( [ inhomog ] ) ] . therefore ,
using the final value theorem @xmath86 @xcite , we find that @xmath87 this means that at long times , the generalized master equation behaves like the markovian master equation @xmath88 defining the rate matrix @xmath89 where @xmath90 we can rewrite ( [ mefake ] ) as @xmath91 .
using the perron - frobenious theorem , all eigenvalues of @xmath92 are negative aside from one which is zero .
the probability @xmath49 therefore decays exponentially to a steady state solution @xmath93 that obeys the condition @xmath94 the steady state solution corresponds to equilibrium if the detailed balance condition @xmath95 is satisfied , that is , if all fluxes between pairs of states associated with the different mechanisms @xmath10 become zero at equilibrium .
this would not be possible if the long - time statistics of the different mechanisms were different .
if the detailed balance condition is not satisfied , then the solution @xmath96 is a nonequilibrium steady - state .
a useful connection to thermodynamics is provided if we consider that each state @xmath9 has a given energy @xmath97 and number of particles @xmath98 and that each different mechanism @xmath10 inducing transitions between states corresponds to a given reservoir with a given temperature @xmath99 and chemical potential @xmath100 .
we then assume that @xmath101 equation ( [ detailbalance ] ) holds only if all the @xmath102s and @xmath100s are equal . in this case
the equilibrium distribution correspond to the grand canonical ensemble .
if the @xmath102s and @xmath100s are different , the steady state is a non - equilibrium steady state that obeys eq .
( [ ness ] ) .
when @xmath103 , we must separately specify the statistical properties of the waiting time distribution for the first jump after @xmath43 .
we choose @xmath104 because other choices add only further complications but little of general interest to our specific problem .
the inhomogeneous initial condition term then drops out , and for small @xmath16 , using ( [ ratebis ] ) , eq .
( [ generalizedmelaplace ] ) becomes @xmath105 using the rules of fractional calculus @xcite , this equation can be written in the time domain as @xmath106 where the riemann - liouville fractional integral is given by @xmath107 if the matrices with elements @xmath108 can be diagonalized and the eigenvalues @xmath109 ( which are all negative or zero ) are non - degenerate , the solution of eq .
( [ generalizedmeanomal ] ) can be written as a linear combination of solutions of @xmath110 the solution of this equation is the mittag - leffler function @xmath111 , where @xmath112 ( see appendix b of ref .
@xcite ) . at long times
the mittag - leffler function decays as a power law , @xmath113 ( at short times it behaves as a stretched exponential , @xmath114 ) .
thus the general solution of the generalized master equation is a linear combination of mittag - leffler functions , and the probability @xmath49 decays toward the zero eigenvalue node as a power law .
although our ultimate goal is to establish conditions under which fluctuation theorems are valid for systems whose dynamics are described by ctrws , we first consider the counting statistics for such a system .
these statistics are interesting because they are experimentally accessible , and this analysis leads to some definitions that are useful in the discussion of fluctuation theorems .
consider a system described by the ctrw of sec .
[ ctrw ] where each microscopic state @xmath9 has a given number of particles @xmath98 and an energy @xmath97 , and where the allowed transitions between pairs of states are due to different mechanisms @xmath10 , each corresponding to an reservoir @xmath10 .
we want to calculate the probability @xmath115 that an energy transfer @xmath116 and a matter transfer @xmath117 occurs between the system and the reservoir @xmath10 during time @xmath14 .
we define the set of parameters @xmath118 and the generating function @xmath119}.\end{aligned}\ ] ] the probability can be recovered from the generating function using @xmath120 derivatives of the generating function with respect to the elements of @xmath121 evaluated at @xmath122 gives the moments of the distribution of this process .
+ we define @xmath123 which is the wtd matrix whose elements associated with a transition caused by mechanism @xmath10 are weighted by the exponential of @xmath124 ( @xmath125 ) times the change of energy ( matter ) that this transition induces .
we note that @xmath126 .
to evaluate the generating function and the associated moments , we replace the wtd @xmath15 by ( [ gfratemat ] ) in the ctrw ( [ ctrw1 ] ) and ( [ ctrw2 ] ) , and thus obtain @xmath127 and @xmath128 by doing so , we are weighting the probability of all the trajectories of length @xmath14 ending up in the state @xmath9 and along which a transfer of energy ( matter ) between the system and the reservoir @xmath10 occurs , by the exponential of @xmath124 ( @xmath125 ) times this energy ( matter ) transfer . by summing over all final states @xmath9 ,
we reconstruct the generating function ( [ probtransferinverse ] ) as @xmath129 where @xmath130 is the unit vector . proceeding as in sec .
[ ctrw ] , the formal solution of ( [ ctrwgf1 ] ) and ( [ ctrwgf2 ] ) in laplace space leads to the general solution for the generating function @xmath131 if ( [ ctrwgfformalsol ] ) can be evaluated and inverse laplace transformed , it provides the full statistics of the energy and matter transfer for a finite time interval .
this is often a difficult task , and one therefore often focuses on the long - time behavior .
this behavior is accessed through the solution ( [ ctrwgfformalsol ] ) and also through the equation of motion for the generating function , which can be deduced by proceeding in the same way as in sec .
[ gme ] when deriving the generalized master equation from ( [ ctrwgf1 ] ) and ( [ ctrwgf2 ] ) . defining @xmath132
we find @xmath133 where @xmath134 to calculate the long - time behavior of the moments of the probability distribution of heat and matter transfer between the system and its reservoir , we first consider the situation when a jump occurred at time zero .
we will comment later on situations in which this is not the case . in the long time
limit , ( [ ctrwgfformalsol ] ) diverges at @xmath122 , because @xmath135 , a limit which is singular because the determinant of @xmath136 is zero .
this can be seen by considering the transpose of the matrix ( transposition does not affect the determinant ) and by replacing its first column by the sum of all the columns of the matrix ( the determinant is not affected by replacing a column by a linear combination of it with other columns ) , which only contains zeros since @xmath137 ( the determinant of a matrix with a zero column is zero ) .
to lowest order in @xmath16 , the determinant behaves like @xmath138 .
this means that at small @xmath16 , @xmath139 . using the long time behavior of the wtd as expressed in ( [ s0wtdpsigeneral ] ) and ( [ s0wtdphigeneral ] ) , and using ( [ ctrwgfformalsol ] )
, we thus see that @xmath140 .
we could have arrived at this directly because @xmath141 implies @xmath142 .
this reasoning makes it clear that in order to calculate moments , the derivatives with respect to one of the @xmath143 s evaluated at @xmath122 must be calculated from ( [ ctrwgfformalsol ] ) before the long time limit is taken . upon taking the @xmath7th derivative of ( [ ctrwgfformalsol ] ) with respect to one of the @xmath143 s at @xmath122 , the dominant contribution at small
@xmath16 is @xmath144 we used the fact that at small @xmath16 , @xmath145 @xmath146 .
@xmath147 is a diagonal matrix with diagonal elements @xmath148 .
using tauberian theorems @xcite , we conclude that the moments of the energy and matter transfer between the system and the reservoir @xmath10 will behave at long times as @xmath149 as noted earlier , these moments are experimentally accessible .
let us now turn back to the case where no jump occurred at time zero .
we only comment on wtds with @xmath40 since only then is it possible to carry out the averaging procedure described in sec .
[ ctrw ] . since @xmath150 and @xmath151 , eq .
( [ momentasymptoticenergymatter ] ) with @xmath1 is still valid , perhaps with a different proportionality factor . in appendix [ qd ]
we explicitly implement these ideas by calculating the moments for a two level quantum dot .
we now explore the conditions under which a fluctuation theorem holds for a ctrw .
two types of derivations have been used to obtain steady state fluctuation theorems for stochastic dynamics .
the first relies on a symmetry of the generating function , which translates into a fluctuation theorem using large deviation theory as in refs .
the other exploits the specific form of the logarithm of the ratio of the probability of a trajectory and the probability of its time reversed trajectory .
we will use both for ctrws and find that they can only be considered equivalent under specific conditions . this approach to arrive at a fluctuation theorem can only be implemented for @xmath1 because it requires a finite mean waiting time between transitions .
we thus restrict our discussion to this case .
we have seen in sec .
[ longtimegme ] that as long as one considers wtds with @xmath1 , the generalized master equation at long times behaves like a markovian master equation and thus reaches a steady state . using the same arguments
, we can show that the equation of motion ( [ generalizedmemod ] ) for the generating function behaves at long times as @xmath152 where @xmath153_{mm ' } \equiv \tilde{w}_{mm'}(\boldsymbol{\gamma},0 ) \ \ { \rm for } \ \ m \neq m ' \nonumber\\ & & [ \boldsymbol{w}(\boldsymbol{\gamma})]_{mm } \equiv - \sum_{n } \tilde{w}_{nm}(0 ) .
\label{defrategf}\end{aligned}\ ] ] this implies that for long times @xmath154 where @xmath155 and where @xmath156 is the dominant eigenvalue of @xmath157 .
this dominant eigenvalue gives the cumulant generating function because @xmath158 note that the derivation to follow does not explicitly require the generalized master equation to be equivalent to a markovian master equation at long times ; what is required is the limiting behavior ( [ cumulantgflong ] ) .
furthermore , using ( [ rateratio1res ] ) , we can verify that ( [ defrategf ] ) satisfies @xmath159 where @xmath160 .
since these two matrices have the same eigenvalues , this implies the symmetry @xmath161 large deviation theory @xcite can now be applied . the limiting behavior ( [ cumulantgflong ] ) and the symmetry ( [ symmetryvp ] ) then imply the fluctuation theorem for the probability @xmath162 for the energy and matter currents ( cf . below ) between the system and the reservoir @xmath10 at long times @xcite , @xmath163 here @xmath164 represents the change of entropy due to the exchange processes with the reservoirs .
we note that the change in energy and matter can be written in terms of energy and matter currents as @xmath165 so that in the long time limit @xmath166 and @xmath167 correspond to steady state currents .
this is why ( [ detailedftcurrent ] ) is called a current fluctuation theorem @xcite .
we now define the matrix @xmath168 as @xmath169 where @xmath121 is replaced by @xmath170 . obviously , when replacing @xmath169 by @xmath168 in our previous results , we calculate the statistics of @xmath171 .
the symmetry ( [ symmonwflux ] ) now implies that @xmath172 so that @xmath173 using again large deviation theory , we get the fluctuation theorem @xmath174 in the steady state , the average energy and matter transferred with a reservoir @xmath10 , can be obtained by taking the derivative with respect to @xmath175 at @xmath122 of the formal solution of ( [ eqmotiongfalphaone ] ) .
we get @xmath176 current conservation at steady state follows from @xmath177 if we assume that all @xmath178 reservoirs ( @xmath179 ) have different temperatures and chemical potentials , we have @xmath180 independent nonequilibrium forces associated to energy and matter transfer that can be defined as @xmath181 where @xmath182 . using ( [ currentconservation ] ) , the average change in the entropy due to exchange processes with the reservoirs @xmath183 can be written in the familiar thermodynamical form of the entropy production in a steady state @xmath184 when @xmath103 , we have seen in section [ longtimegme2 ] that the solution of the gme behaves at long times as a power law @xmath185 .
this will also be the case for the solution of the equation of motion for the generating function ( [ generalizedmemod ] ) .
therefore , contrary to ( [ cumulantgflong ] ) , the cumulant generating function @xmath186 is zero .
this indicates that the cumulants decay slower than @xmath14 .
the fluctuation theorem symmetry [ symmertryvp ] is only present in the eigenvalues of the generator of the gme ( [ eqmotiongfalphaone ] ) but not in the eigenvectors . for @xmath1 , the cumulant generating function
is given by the dominant eigenvalue , so that the fluctuation theorem will reflect itself on all quantities relates to it ( large deviation theory is precisely used to make the link between this generating function and the probabilities ) .
however , in the case @xmath103 , there is no way to separate in the long time limit , the contribution of the eigenvalues from the contribution of the eigenvectors to the statistics thus preventing a fluctuation theorem to hold .
we denote a forward trajectory of the system between times @xmath43 and @xmath187 by @xmath188 . as illustrated in fig .
[ figtraj ] , at @xmath43 the system is in state @xmath189 and stays there until it jumps to state @xmath190 at time @xmath191 via mechanism @xmath192 .
it remains there until time @xmath193 , when it jumps to state @xmath194 via mechanism @xmath195 .
the trajectory continues in this fashion ; at time @xmath196 there is a jump to state @xmath197 , where the system remains at least until time @xmath198 .
the total number of jumps in this trajectory is @xmath199 .
the probability of this trajectory is @xmath200=\rho_{m_0}(0 ) \psi_{m_1 m_0}'^{(\nu_1)}(\tau_1 ) \nonumber \\ & & \hspace{1.5cm}\times \bigg ( \prod_{i=1}^{n-1 } \psi_{m_{i+1}m_{i}}^{(\nu_{i+1 } ) } ( \tau_{i+1}-\tau_{i } ) \bigg ) \phi_{m_{n}}(t-\tau_n ) .
\nonumber \\ & & \end{aligned}\ ] ] the time - reversed trajectory @xmath201 starts in state @xmath197 at time @xmath198 , jumps to state @xmath202 at time @xmath203 via mechanism @xmath204 , and so on . at time
@xmath205 a jump to state @xmath190 occurs via mechanism @xmath192 , and the system remains there until at least time @xmath206 .
the probability of this time reversed trajectory is @xmath207 = \rho_{m_n}(t ) \psi_{m_{n-1 } m_{n}}'^{(\nu_{n-1})}(t-\tau_n ) \nonumber \\ & & \hspace{1.5cm}\times \bigg ( \prod_{i=0}^{n-2 } \psi_{m_{i}m_{i+1}}^{(\nu_{i } ) } ( \tau_{i+2}-\tau_{i+1 } ) \bigg ) \phi_{m_{0}}(\tau_1 ) .
\nonumber \\ & & \end{aligned}\ ] ] next we consider the quantity @xmath208 \equiv
\ln \frac{p[m_{\tau}]}{p[\bar{m}_{\tau } ] } , \label{entropyprod}\ ] ] whose explicit form reads @xmath209 & = & \ln \frac{\rho_{m_0}(0)}{\rho_{m_n}(t ) } + \ln \prod_{i=1}^{n-1 } \frac{\psi_{m_{i+1}m_{i}}^{(\nu_{i+1})}(\tau_{i+1}-\tau_{i } ) } { \psi_{m_{i-1}m_{i}}^{(\nu_{i})}(\tau_{i+1}-\tau_{i } ) } \nonumber\\ & & \hspace{0 cm } + \ln \frac{\psi_{m_1 m_0}'^{(\nu_1)}(\tau_1 ) \phi_{m_{n}}(t-\tau_n)}{\psi_{m_{n-1 } m_{n}}'^{(\nu_{n})}(t-\tau_n ) \phi_{m_{0}}(\tau_1 ) } \;. \label{rgeneral } \end{aligned}\ ] ] because each @xmath201 is the unique mirror image of @xmath188 , a sum over all possible forward trajectories is equivalent to a sum over all possible time - reversed trajectories .
therefore , since normalization implies that @xmath210=1 $ ] , an integral fluctuation theorem follows immediately from the definition ( [ entropyprod ] ) , @xmath211 } p[m_{\tau } ] = 1 .
\label{integralft}\ ] ] the positivity of the ensemble average of @xmath212 $ ] , @xmath213 , then follows from jensen s inequality . using the important property @xmath212=-r[\bar{m}_{\tau}]$ ] , which follows form ( [ entropyprod ] ) together with the fact that by taking twice the time - reversal of a trajectory we get back to the original trajectory @xmath214 , we can also derive a detailed fluctuation theorem for @xmath5 , @xmath215 ) p[m_{\tau } ] \nonumber \\ & = & \sum_{m_{\tau } } \delta(r - r[m_{\tau } ] ) { { \rm e}}^{r[m_{\tau } ] } p[\bar{m}_{\tau } ] \nonumber \\ & = & { { \rm e}}^{r } \sum_{\bar{m}_{\tau } } \delta(r - r[m_{\tau } ] ) p[\bar{m}_{\tau } ] \nonumber \\ & = & { { \rm e}}^{r } \sum_{\bar{m}_{\tau } } \delta(r+r[\bar{m}_{\tau } ] ) p[\bar{m}_{\tau } ] \nonumber \\ & = & { { \rm e}}^{r } p(-r ) \label{detailedft } \;.\end{aligned}\ ] ] the integral fluctuation theorem ( [ integralft ] ) and the detailed fluctuation theorem ( [ detailedft ] ) for r are thus completely general and valid for any continuous time random walk .
it is worth mentioning that a similar derivation can be done for any dynamics as long as each trajectory has a corresponding time - reversed trajectory with a nonzero probability @xcite . to make these fluctuation theorems useful , we need to give a physical interpretation to @xmath212 $ ]
. in particular , we will argue that @xmath5 can only be interpreted as a change of entropy if two conditions are satisfied .
one is that the wtds have a finite first moment ( @xmath1 ) .
the other is that they be separable @xcite , that is , that the waiting time distributions can be written as the product of a waiting time portion that depends only on the originating state , and a transition matrix that connects given initial and final states , @xmath216 the separability condition is called direction time independence " by qian and wang @xcite ) . the first term of eq .
( [ rgeneral ] ) , @xmath217 \equiv \ln \frac{\rho_{m_0}(0)}{\rho_{m_n}(t ) } = s_{m_n}(t ) - s_{m_0}(0 ) , \label{trajentropygibbs}\ ] ] where @xmath218 , can be interpreted as a change of system ( gibbs ) entropy along the trajectory because it depends only on the initial and final microscopic states of the system for that trajectory , and the average over trajectories is then simply @xmath219 , where @xmath220 is just a straightforward average over states . in order to interpret the two remaining terms in ( [ rgeneral ] ) , we implement separability of the wtds so that @xmath209 & = & \ln \frac{\rho_{m_0}(0)}{\rho_{m_n}(t ) } + \ln \bigg ( \prod_{i=0}^{n-1 } \frac{p_{m_{i+1}m_{i}}^{(\nu_{i+1 } ) } } { p_{m_{i}m_{i+1}}^{(\nu_{i+1 } ) } } \bigg ) \nonumber\\ & & + \ln \frac{\psi_{m_0}'(\tau_1 ) \phi_{m_{n}}(t-\tau_n ) } { \psi_{m_{n}}'(t-\tau_n ) \phi_{m_{0}}(\tau_1 ) } \;. \label{ratioprobexpli } \end{aligned}\ ] ] now only the wtds of the first jumps remain . in eq .
( [ detailedft ] ) , the path summation runs over all possible trajectories including those with or without a jump at time zero .
one way to handle the problem of having to treat the first jump differently from the others is via the time averaging procedure ( [ wtfirstjump ] ) in ( [ forwardprob ] ) as well as in ( [ backwardprob ] ) .
this can only be done if the first moments are finite .
the third term in ( [ ratioprobexpli ] ) now becomes equal to @xmath221 , so that using ( [ sumratemarkov ] ) the second and third terms in ( [ ratioprobexpli ] ) can be combined and @xmath222 $ ] can be written in terms of the transition matrix elements as @xmath209 = \ln \frac{\rho_{m_0}(0)}{\rho_{m_n}(t ) } + \ln \bigg ( \prod_{i=0}^{n-1 } \frac{\tilde{w}_{m_{i+1}m_{i}}^{(\nu_{i+1})}(0 ) } { \tilde{w}_{m_{i}m_{i+1}}^{(\nu_{i+1})}(0 ) } \bigg ) \;.
\label{formestandard}\end{aligned}\ ] ] this form of @xmath212 $ ] is now exactly the same as the one derived for the markovian master equation ( [ mefake ] ) ( see @xcite ) .
this is a manifestation of the so - called corresponding markov process of a ctrw @xcite .
we denote the second term on the right hand side of ( [ formestandard ] ) as @xmath223 $ ] and call it the reservoir part of the trajectory entropy because it can be interpreted as the change in entropy along the trajectory due to exchange processes between the system and the reservoirs . indeed , using ( [ rateratio1res ] ) , this term can be expressed as @xmath224 & \equiv & \ln \bigg ( \prod_{i=0}^{n-1 } \frac{\tilde{w}_{m_{i+1}m_{i}}^{(\nu_{i+1})}(0 ) } { \tilde{w}_{m_{i}m_{i+1}}^{(\nu_{i+1})}(0 ) } \bigg ) \nonumber\\ & & \hspace{-1.5cm}=\sum_{\nu } \bigg ( -\beta_{\nu } \delta e^{(\nu)}[m_{\tau } ] + \beta_{\nu } \mu_{\nu } \delta n^{(\nu)}[m_{\tau } ] \bigg ) , \label{to_be_labeled}\end{aligned}\ ] ] where the change of energy and number of particles along the trajectory due to the mechanism @xmath10 can be expressed in terms of heat and matter currents along the trajectory as @xmath225 & \equiv & \int_{0}^{t } d\tau i_e^{(\nu)}[m_{\tau } ] \\ \delta n^{(\nu)}[m_{\tau } ] & \equiv & \int_{0}^{t } d\tau i_m^{(\nu)}[m_{\tau}]\end{aligned}\ ] ] [ cf .
( [ ec ] ) and ( [ nc ] ) ] .
these currents are sequences of @xmath226 functions centered at the times of the jumps and multiplied by the corresponding energy or matter change .
they are positive ( negative ) if energy or matter increases ( decreses ) in the system .
we have thus arrived at the important result that for separable wtds with @xmath1 , @xmath212=\delta s[m_{\tau}]+\delta s_r[m_{\tau}]$ ] . since @xmath227 $ ] and @xmath223 $ ]
are interpreted respectively as the change in the system entropy along the trajectory and the change in entropy due to the exchange processes between the system and its reservoirs , it is natural to interpret @xmath212 $ ] as the total change in entropy along the trajectory ( also called the total change in the trajectory entropy production ) @xcite .
note that in this case the principle of microreversibility , implying that at equilibrium the probability of a forward trajectory is identical to the probability of its time reversed trajectory [ @xmath212=0 $ ] ] , is satisfied if the detailed balance condition ( [ detailbalance ] ) is satisfied and if the system is initially at equilibrium [ so that @xmath228 and @xmath229 .
this result is consistent with the findings of ref .
@xcite stating that separability of the wtds and detailed balance are sufficient conditions for microreversibility to be satisfied in a ctrw . in sec .
[ ssfluctheo ] we showed that a fluctuation theorem for @xmath171 can be derived for long times [ see eq .
( [ detailedftres ] ) ] .
in fact , considering separable wtds with @xmath1 , the fluctuation theorem ( [ detailedftres ] ) can be seen as resulting from the fluctuation theorem ( [ detailedft ] ) , as follows .
the quantity @xmath227 $ ] is a bounded quantity which only depends on the probability distribution of the initial and final states of the trajectory . on the contrary ,
@xmath223 $ ] changes each time a jump occurs along the system trajectory .
it is therefore reasonable to assume that for very long trajectories and for the huge majority of realizations , the contribution from @xmath223 $ ] in @xmath212=\delta s[m_{\tau } ] + \delta s_r[m_{\tau}]$ ] will be significantly dominant so that in the long time limit ( [ detailedft ] ) reduces to ( [ detailedftres ] ) .
however , the derivation of ( [ detailedftres ] ) only required wtds with @xmath1 , but did not require separability .
the reason for this requirement in order to identify eq .
( [ detailedft ] ) as a useful fluctuation theorem for all times is that otherwise @xmath212 $ ] has no clear physical interpretation .
the problem is that @xmath212 $ ] ( and , more specifically , the second term in ( [ rgeneral ] ) , which according to our previous argument dominates at long times ) depends on the time intervals between the jumps along the trajectory .
this implies that at the level of a single trajectory , @xmath212 $ ] can not be expressed in terms of exchange processes with the reservoirs ( more precisely , in terms of time integrated currents ) .
however , the fluctuation theorem ( [ detailedftres ] ) indicates that at long times the probability @xmath230 to observe a trajectory such that @xmath212=r$ ] becomes equivalent to the probability to observe a trajectory with a set of time integrated currents so that , via ( [ ftcurrents ] ) , @xmath231 .
one way to understand this result is to coarse - grain the trajectories in time . instead of specifying exactly the time
at which each jump occurs , we define small time intervals of equal size @xmath232 ( sufficiently small so that the probability of observing two transitions in one is interval negligible , i.e. , small compared to the mean time for a transition to occur ) .
we then specify whether a transition between two states occurred or not in this interval . in this way , we define coarse - grained trajectories ( denoted by @xmath233 ) , and we note that different microscopic trajectories can lead to the same coarse grained trajectory . to calculate the probabilities of these trajectories
, we use the fact that at long times the dynamics is described by ( [ mefake ] ) , which can be discretized in time intervals @xmath232 .
this discretized form allows us to identify the probability to stay in a given state @xmath9 during a given time interval with @xmath234 , and the probability to jump from @xmath9 to @xmath13 by mechanism @xmath10 with @xmath235 .
using these probabilities , we can construct the probability of a coarse - grained trajectory and that of its time reversed coarse - grained trajectory .
the logarithm of their ratio gives @xmath236+\delta s_r[\tilde{m}_{\tau}]$ ] , with the same definitions as in ( [ trajentropygibbs ] ) and ( [ to_be_labeled ] ) , but for @xmath233 instead of @xmath188 . with the argument that at long times @xmath237 $ ] will be dominant ,
we understand why the fluctuation theorem ( [ detailedftres ] ) holds at long times .
one should note that long times are also needed in this case to get rid of the contribution from the initial part of the trajectory where ( [ mefake ] ) is not valid .
also , contrary to the case where separability holds , it is only at the coarse grained level and considering very long trajectories that such a dynamics can satisfy `` microscopic '' reversibility .
the fluctuation theorem ( [ detailedftres ] ) does not result from a probabilistic asymmetry under time reversal at the microscopic trajectory level as in the case of separable wtds , but at a coarse grained level and only for long times .
we can finally comment on the case of wtds with diverging mean waiting times . even if separability is satisfied , because mean waiting times diverge , ( [ rateratio1res ] ) can not be used to express @xmath212 $ ] ( or parts of it ) in term of exchange processes with the reservoirs . the fluctuation theorem ( [ detailedft ] )
is thus valid but seemingly useless .
in this paper we have considered continuous time random walks ( ctrws ) in which multiple mechanisms that may have different statistical properties induce transitions between pairs of states . a given energy and number of particles
can be attributed to each state , so that each mechanism can be thought of as corresponding to a given reservoir .
if these reservoirs have different statistical properties , such a ctrw therefore describes the stochastic dynamics of an open system .
the statistics of the transitions between states associated with each mechanism are described by waiting time distributions ( wtds ) .
if a wtd decays at long times at least as fast as @xmath2 , the distribution has a finite first moment and we say that @xmath1 . if the decay is slower , as @xmath4 where @xmath6 , all moments diverge .
we have analyzed the long - time behavior of the probability @xmath49 that the system is in state @xmath9 at time @xmath14 via the generalized master equation for this probability .
when @xmath1 the generalized master equation leads at long times to an exponential decay of the probability to a steady state , exactly as it would for an ordinary markovian master equation .
if the wtds corresponding to the different mechanisms are different , the steady state is a nonequilibrium steady state .
if they are identical , the steady state distribution corresponds to equilibrium and satisfies detailed balance . on the other hand ,
if @xmath6 , the probability @xmath49 evolves as a linear combination of power law decays @xmath238 towards the zero eigenvalue mode , and the system never reaches a steady state .
we have presented a formalism to calculate the counting statistics of the energy and matter transfer in the ctrw based on a generating function propagation method . by considering systems exchanging energy and matter with different reservoirs
, we have shown that @xmath7th moment of the probability distribution to exchange a certain amount of energy or matter between the system and a given reservoir during a time interval @xmath14 behaves as @xmath8 at long times .
this result holds for @xmath239 and reflects the subordination principle @xcite . using our generating function formalism together with large deviation theory , for wtds with @xmath1 we derived a fluctuation theorem for the trajectory quantity @xmath223 $ ] representing the change of entropy along the trajectory due to exchange processes with the reservoirs [ which can be explicitly related to the energy and matter currents between the system and the reservoirs ( [ ftcurrents ] ) ] .
if @xmath240 is the probability to observe a trajectory along which occurs a change @xmath223=\delta s_r$ ] , the fluctuation theorem reads @xmath241 and is only valid at long times . for wtds with diverging first moments , @xmath242
, this fluctuation theorem does not hold .
the trajectory quantity @xmath212 $ ] is defined as the logarithm of the ratio of the probability of a given trajectory to the time reversed trajectory .
we have shown that for any ctrws , the ensemble average of @xmath212 $ ] is always positive and a fluctuation theorem stating that the probability @xmath230 to observe a trajectory such that @xmath212=r$ ] is exponentially more likely than to observe a trajectory such that @xmath212=-r$ ] , i.e. , @xmath243 always holds .
separable wtds are ones for which the state - directional part and the temporal part of the wtd factorize [ see ( [ dti ] ) ] .
we have shown that it is only for separable wtds with @xmath1 that @xmath212 $ ] can be interpreted as the change of entropy along the trajectory . in this case , this fluctuation theorem is related to the previous one because @xmath212=\delta s[m_{\tau}]+\delta s_r[m_{\tau}]$ ] , where @xmath227 $ ] is the change of system entropy along the trajectory and is bounded , contrary to @xmath223 $ ] which typically grows for long trajectories . therefore , for most realizations @xmath212 \approx \delta s_r[m_{\tau}]$ ] at long times , thus providing an interpretation of the fluctuation theorem for @xmath223 $ ] in terms of probabilistic asymmetry under time reversal of the probability of a trajectory . at equilibrium ,
the symmetry is restored and @xmath212=0 $ ] as expected from the principle of microreversibility . for nonseparable wtds with @xmath1 ,
@xmath212 $ ] can only be related to a change of entropy after a coarse graining of the trajectories in time .
the fluctuation theorem for @xmath223 $ ] still holds , but can only be interpreted as a measure of a probabilistic asymmetry in time at the level of coarse grained trajectory ( and also only for long times ) , and not at the level of the microscopic trajectories as in the case of separable wtds .
m. e. would like to thank frantiek anda and shaul mukamel for early discussions which motivated this work .
m. e. is supported by the fnrs belgium ( charg de recherches ) and by the luxemburgish government ( bourse de formation recherches ) .
this research is supported in part by the national science foundation under grant phy-0354937 .
as an application of the counting statistics results of sec . [ gfcurrentsec ] , we consider a single level quantum dot between left and a right leads @xmath244 .
there are two states in the system , one corresponding to the empty level @xmath245 and the other to the filled level @xmath246 .
since the energy transfer between the system and a reservoir is directly proportional to the particle transfer in this model , we only consider particle transfer .
we take separable wtds , specifically @xmath247 where @xmath248 note that @xmath249 we define the affinities @xmath250 , where @xmath251 , and @xmath252 .
+ from now on , we focus on the net particle transfer between the left lead and the dot so that @xmath253 and @xmath254 . we also note that @xmath255 the generating function ( [ ctrwgfformalsol ] ) for this model reads the wtds are taken to be of the form @xmath261 where @xmath262 . if @xmath1 , @xmath263 and @xmath264 .
+ we can now confirm that if @xmath6 and a jump occurred at time zero ( @xmath265 ) , or if @xmath1 and a jump occurred at time zero ( @xmath265 ) , or if @xmath1 and we do nt know at time zero when the last jump occurred ( @xmath266 ) , we always get @xmath267 where @xmath268 . using the tauberian theorem ,
this leads to @xmath269 when @xmath1 , ( [ current ] ) can be written as @xmath270 where @xmath271 and @xmath272 .
this is the same result as obtained in ref
. @xcite using a markovian master equation . in
can easily be seen that ( [ currentbis ] ) vanishes when detailed balance is satisfied . finally , we note that the cumulants calculated from these moments all vanish .
this indicates that we should calculate not only the leading time dependence of the moments as we do when applying tauberian theorems , but must retain higher orders to extract information about the time dependence of the cumulants .
b. d. hughes , _ random walks and random environments : volume 1 : random walks _ , ( oxford university press , usa , 1995 ) ; _ random walks and random environments : volume 2 : random environments _ ( oxford university press , usa , 1996 ) . m. m. meerschaert , d. a. benson , h .- p .
scheffler and b. baeumer , phys .
e * 65 * , 041103 ( 2002 ) ; a. a. satinislavsky , physica a * 318 * , 469 ( 2003 ) ; r. gorenflo , f. mainardi and a. vivoli , chaos , sol .
frac . * 34 * , 89 ( 2007 ) . | we consider continuous time random walks ( ctrw ) for open systems that exchange energy and matter with multiple reservoirs . each waiting time distribution ( wtd ) for times between steps
is characterized by a positive parameter @xmath0 , which is set to @xmath1 if it decays at least as fast as @xmath2 at long times and therefore has a finite first moment .
a wtd with @xmath3 decays as @xmath4 .
a fluctuation theorem for the trajectory quantity @xmath5 , defined as the logarithm of the ratio of the probability of a trajectory and the probability of the time reversed trajectory , holds for any ctrw .
however , @xmath5 can be identified as a trajectory entropy change only if the wtds have @xmath1 and satisfy separability ( also called direction time independence " ) . for nonseparable wtds with @xmath1 , @xmath5
can only be identified as a trajectory entropy change at long times , and a fluctuation theorem for the entropy change then only holds at long times . for wtds with @xmath6
no meaningful fluctuation theorem can be derived .
we also show that the ( experimentally accessible ) @xmath7th moments of the energy and matter transfers between the system and a given reservoir grow as @xmath8 at long times . |
light rays from distant sources carry unique information about the matter and gravitational potential along the line - of - sight .
a well - known example is the signature of intervening gas clouds imprinted into spectra of background sources via absorption lines .
mass concentrations located along the path of photons can also induce gravitational lensing effects .
background sources can be magnified , as detected by @xcite , and galaxy shapes can be distorted as measured through galaxy - galaxy lensing ( cf .
@xcite , @xcite , @xcite , @xcite ) and cosmic shear ( e.g. ) .
measuring these effects has become a powerful tool for probing the mass distribution in the universe .
in addition , dust extinction effects are expected to occur as radiation pressure from stars and galactic winds triggered by supernovae are expected to expel some amount of dust from galactic disks into the intergalactic medium @xcite .
detecting dust reddening by galaxy halos would provide us with useful information on the life cycles of dust particles as well as characterize the opacity of the universe . in practice , detecting such an effect is made difficult by the requirement to measure brightness and/or color changes at a sub - percent level on 100 kpc scales around galaxies . a first attempt to find dust in galactic halos
was made by @xcite who reported a 3-@xmath10 indication for a color change of background galaxies found around two nearby spiral galaxies . probing dust reddening induced by galaxy halos has not been revisited since then , despite the dramatic improvement in data quality and sample size . in this work
we investigate simultaneously gravitational lensing and dust reddening of background quasars by foreground galaxies and associated large scale structure .
our observational results primarily make use of the angular cross - correlation between the brightness of quasars and the projected density of galaxies .
we first recover and improve upon previous measurements of the magnification of quasar images by gravitational lensing due to galaxies located nearby the light paths using a sample of 85,000 quasars behind 20 million galaxies derived from the sloan digital sky survey ( @xcite ; sdss ) data release 3 ( @xcite ) .
second , this large sample together with high accuracy photometry in five optical pass bands allows us to detect the presence of dust in the intervening space and explore its distribution and properties .
this allows us to study the properties of intergalactic dust and provides a way of inferring the abundance of dust in the universe .
+ we introduce the formalism of brightness - density correlations in 2 .
data processing and measurements are presented in 3 & 4 .
the astrophysical results are given in 5 and we summarize in 6 .
when needed we use @xmath11 , @xmath12 and @xmath13 .
let us begin by considering a background source angularly close to a foreground galaxy .
this galaxy acts as a gravitational lens , magnifying the source flux and giving rise to extinction and reddening due to the presence of dust surrounding it .
the observed flux is then modified by the combination of these effects according to = f_0e^- _ , where @xmath14 is the gravitational magnification and @xmath15 is the optical depth for dust extinction at an observed wavelength @xmath16 . the corresponding magnitude shift is m_=-2.5+_. + [ eq : main ] when @xmath14 departs weakly from unity , we can re - express this relation as @xmath17 where @xmath18 . thus , magnification and extinction compete in changing the brightness of background sources .
dust extinction is in general wavelength dependent while magnification is achromatic , so the two effects can , in principle , be separated using multi - color data .
below we show how correlations between the density of foreground galaxies and the magnitude of background sources allow us to constrain the galaxy - mas and galaxy - dust correlation functions .
the galaxy - mass correlation is a powerful tool to probe the connection between matter and galaxies , providing direct constraints on the dark matter distribution and models of galaxy formation . to introduce it
, we first define the galaxy and mass overdensities : _
g(*x*)=-1 and _ m(*x*)=-1 where @xmath19 and @xmath20 are the density of galaxies and matter .
the galaxy - mass correlation is then defined by @xmath21 this cross - correlation can be related to the projected surface density of galaxies : @xmath22 the first relation indicates that the galaxy - mass correlation is equal to the mean mass profile around galaxies , at a given separation @xmath23 .
the second relation is simply a projection of the 3-dimentional galaxy - mass correlation introduced above in equation [ eq : gm_3d ] and where @xmath24 .
the mass surface density @xmath25 can be probed with gravitational lensing . in the weak lensing regime , it is straightforwardly related to the observable magnification , according to @xmath26 here @xmath27 is the lensing convergence and the critical mass surface density is given by @xmath28 where @xmath29 and @xmath30 are respectively the angular diameter distances to the lens , the source and between the lens and the source . as indicated in equation [ eq : delta_m ] , magnification will affect the brightness of background sources and induce a correlation between the density of foreground galaxies and the magnitude of background sources . in order to understand the impact on observable quantities ,
let us consider a given area of the sky and let @xmath31 be the intrinsic magnitude distribution of some sources .
the photons originating from these sources may be deflected by gravitational lensing and magnification affects their magnitude distribution such that @xmath32 where @xmath33 is the induced magnitude shift .
this leads to an observable mean magnitude shift : @xmath34 it should be noted that for a magnitude - limited sample of sources , the mean magnitude shift induced by a population of foreground galaxies , @xmath35 , differs from the observable mean magnitude shift of the individual sources , @xmath36 .
the difference between the two depends on the shape of the source magnitude distribution @xmath37 and the limiting magnitude @xmath38 of the sample . in the case where the induced magnitude shift @xmath39 is small
compared to the limiting magnitude of the sample , the difference between the observed and induced magnitude shift can be linearized in @xmath35 and we have @xmath40 where the coefficient @xmath41 depends on the shape of the magnitude distribution and the limiting magnitude . in the considered limit , it is given by @xmath42 \label{eq : cs}\ ] ] ( see the derivation in the appendix ) .
if the sample of sources is not magnitude - limited , @xmath43 and @xmath44 , i.e. the measured magnitude shift equals the induced one . for magnitude - limited samples , in general we have @xmath45 , i.e. the observable magnitude shift is smaller than the induced one .
it is important to note that for power - law luminosity distributions @xmath46 .
therefore , while the brightness of each object changes by @xmath39 , the mean magnitude of the sample remains unchanged .
this is due to the inclusion of sources which become brighter than the limiting magnitude . for sdss quasars in the @xmath47-band , down to a limiting magnitude of @xmath48
, we find @xmath49
. + let us suppose that a population of background quasars and foreground galaxies are sufficiently well separated in redshift so that physical correlations between them can be neglected .
gravitational lensing will give rise to an apparent correlation between quasar brightness and the proximity to a galaxy overdensity , which we can write as @xmath50 where @xmath51 is the quasar magnitude fluctuation and @xmath52 is the foreground galaxy overdensity at a given angular distance @xmath23 from a background source . if magnification effects are in the weak regime , i.e. @xmath53 , the above relation reads @xmath54 on scales below a few arcminutes , a non - linear treatment of the magnification as a function of the density contrast is needed .
such calculations are presented in , where the authors show that non - linear effects can increase the amplitude of the magnification by about 30% at a scale of one arcminute . for populations of quasars and galaxies selected over an appreciable redshift range , the above correlation can be expressed in terms of the galaxy - dark matter cross - power spectrum using the limbers approximation . following the formalism introduced by and the notation laid out in @xcite ,
we can write @xmath55 where @xmath56 is the cosmological matter density relative to critical , @xmath57 is the comoving distance , @xmath58 is the lensing kernel , and @xmath59 is the galaxy - dark matter cross - power spectrum .
the quantity @xmath60 therefore probes the magnification due to individual galaxy halos on small scales and on large scales it constrains the large scale distribution of matter in the universe .
+ statistical properties of the distribution of dust around galaxies can be constrained by the galaxy - dust correlation function : @xmath61 where @xmath62 is the dust density fluctuation .
this cross - correlation can be related to the projected dust surface density of galaxies : @xmath63 at optical wavelengths , dust extinguishes and reddens the light of background sources .
the galaxy - dust correlation can then be probed by measuring the cross - correlation between the colors of background sources and the distribution of foreground matter .
let us define the magnitude shift of a background source measured at a wavelength @xmath64 with respect to the mean magnitude of the sample , m_(**)=m _ ( * * ) - m_. expanding the formalism introduced in the previous section to dust reddening effects , and defining the color excess @xmath65
\end{aligned}\ ] ] we introduce the quasar color - galaxy density correlation , which can be used to probe the galaxy - dust correlation : @xmath66 \,\rangle(\theta)\;. \label{eq : edelta}\ ] ] if the foreground galaxy and background source populations are well separated in redshift , the above correlation provides us with information on the mean reddening and therefore the mean amount dust around galaxies .
the information at different wavelengths can be used to probe the shape of the extinction curve . for scales on the order of a typical galaxy virial radius and smaller
, the signal is expected to be dominated by single galaxies and provides with the average dust density profile around galaxies . on larger scales
it provides information as to the large scale distribution of dust in the universe . for a given extinction curve
the above quantity can be used to infer the mean dust extinction profile around galaxies @xmath67 , or similarly the mean optical depth for dust extinction @xmath68 .
it should be noted that not only extinction but also reddening effects can cause quasars to drop out of the selection criteria as their identification is a function of colors . in the latter case
, the measured reddening excess turns out to be lower than the true value and the relation between observed and induced color change needs to be quantified . in the present analysis ,
reddening effects are sufficiently small for this to be negligible ( see discussion below ) .
+ the measurement of these galaxy - mass ( equation [ eq : gm_2d ] ) and galaxy - dust ( equation [ eq : edelta ] ) correlations will be the focus of the remainder of this paper .
the data set consists of galaxy and quasar catalogues and is drawn from the third sdss data release ( dr3 ; @xcite ) .
the survey provides images in five broad optical bands ( @xmath69 @xcite ) . before masking ,
this set covers roughly 5300 square degrees , the majority of which is located around the north galactic cap . to reduce systematic errors in the photometric data , we impose a seeing limit of 14 and a galactic extinction limit of 0.2 in the @xmath70 band ( @xcite ) .
we also apply a mask blocking a one - arcminute radius around bright galaxies ( @xmath71 ) and stars with saturated centers to prevent changes in the source number densities due to local fluctuations induced by errors in sky brightness subtraction ( @xcite ) .
altogether , the masks reduce our total area to @xmath72 square degrees . with these cuts , we can reliably perform star / galaxy separation using bayesian methods to @xmath73 ( @xcite ) , yielding a galaxy sample whose density is independent of local variations in seeing , galactic extinction , sky brightness or stellar density after masking .
this yields about 24 million galaxies between @xmath74 at a density of approximately 1.8 galaxies per square arcminute .
the @xcite paper used a sample that was instead r - band limited , which accounts for the difference in the galaxy sample sizes . for galaxy magnitudes , we use composite model magnitudes , constructed from the _ expmag
_ , _ devmag _ and _ fracdev _ parameters in the sdss database .
this provides a more robust estimate of galaxy flux at faint magnitudes than the petrosian magnitudes used in the sdss spectroscopic sample , which can have strong variations due to local seeing for @xmath75 . for quasars ,
we consider the flux within the psf profile , designated as _
psfmag_. all magnitudes are de - reddened to correct for galactic extinction before applying the various magnitude cuts .
the quasar data set was generated using the kernel density estimation ( kde ) method described in @xcite applied to the dr3 data set .
the kde method is an extension of the traditional color selection technique for identifying quasars .
two training sets , one for stars and one for quasars are prepared and the colors for each object are compared to those of the two training sets using a 4d euclidean distance .
the objects are then classified as either quasar or star according to a larger probability of membership .
this technique allows a clean separation of relatively low redshift ( @xmath76 ) quasars from the stellar locus , producing a catalog of 225,000 quasars down to a limiting magnitude of @xmath48 with efficiency and completeness greater than the sdss spectroscopic target selction algorithm ( @xcite ) . after masking ,
the total quasar population is reduced to 195,000 .
in addition to finding quasars , we applied photometric redshift techniques ( @xcite ) to exclude low redshift quasars which might be physically associated with our foreground sample .
quasar photometric redshifts are driven by the broad spectral emission features in their spectra , resulting in photometric redshift likelihoods that can have multiple peaks as well as strong asymmetries around the most likely redshift . rather than estimating a gaussian redshift error , we use an upper and lower redshift bounds for a specified likelihood . to minimize the overlap with the galaxies in the redshift space , we require the upper and lower bounds to be within the range @xmath77 . the corresponding distribution of quasar redshift probabilities is shown in figure [ fig : redshift_distribution ] with the red line . for simplicity
we treat the redshift p.d.f . of each quasar as a flat distribution between the upper and lower redshift bounds described above and weighted by the likelihood that the redshift was within those bounds .
this final selection criterion reduces the number of quasars to 85704 .
we estimate the overall shape of the galaxy redshift distribution based on the cnoc2 luminosity functions ( @xcite ) following the treatment described in @xcite .
this distribution is well fit by the expression ( ) _ gz^2e^-(z /0.187)^1.26 , [ eq : redshift_gal ] plotted with the solid blue curve . the mean redshift for the galaxy sample is found to be @xmath78 for lensing purposes it is useful to compute the galaxy redshift distribution weighted by the expected lensing efficiency .
we have computed this quantity using the quasar redshift distribution and we show it using the blue dashed line in the figure . the effective redshift weighted by the lensing efficiency is given by @xmath79 it is found to be @xmath80 .
figure [ fig : magnitude_g ] shows the quasar magnitude and color distributions . as can be seen , the quasar sample is magnitude limited in @xmath47 but not color limited . while quasars are selected both in magnitude and color space , the modes of the color distributions are far from the limiting colors . as a result , a measured color change can be directly used as an estimate of the intrinsic color change of the population .
such a property is not satisfied for magnitude changes ( see section [ sec : extinction ] for more details ) .
we measure the density of galaxies ( taking into account missing area due cuts on the local seeing variations , bright stars , etc . ) and compute its correlation , @xmath81 , with the magnitude of background quasars in the band @xmath82 , as a function of angular separation : @xmath83 on small scales ( @xmath84 ) , the signal contributions from magnification and dust extinction are expected to be dominated by poisson noise . for larger scales , measurements from different angular bins
become significantly correlated , as one would expect since different quasars will be sampling the same local population of galaxies .
since we are looking for very small variations in the quasar magnitude and galaxy density , the photometric calibration across the survey needs to be highly homogeneous . to avoid possible drift in the sdss calibration over time
, we measure the qso - brightness galaxy - overdensity correlation separately on each stripe of the survey , and then take the average over all stripe - based estimators : w_j ( ) = _
i=1^n_stripe w_^(i ) ( ) , where @xmath85 is the qso - magnitude galaxy - overdensity correlation measured for stripe @xmath86 using the filter @xmath82 .
the associated cost is to reduce the sensitivity of the estimator to power arising from one direction , i.e. , the scanning direction , which is equivalent to reducing the size of the sample when measuring the large - scale power .
our strongest signal is on angular scales smaller than the width of a single stripe ( @xmath87 ) so the extent of this effect should be minimal . + in order to measure reddening induced correlations , we compute the difference between two stripe - based estimators at different wavelengths , w _ ( ) = w _ ( ) - w _ ( ) . to estimate the error on the above estimators we use @xmath88 bootstrap samples of the quasar catalog with which we re - measure the above quantities and estimate their dispersion .
following the above procedure we measure the correlation between the observed qso brightness excess and galaxy density @xmath89 ( equation [ eq : w_definition ] ) , where @xmath82 denotes one of the five sdss pass bands , with the errors estimated from bootstrap resampling .
the results shown in figure [ fig : mag_shift_log ] are examples for the @xmath47 and @xmath86 bands .
we observe that quasar magnitude shifts and galaxy overdensities are anticorrelated .
in other words , quasars appear to be brighter when closer to galaxies , which implies a dominance of magnification over extinction effects for the main sample of sdss galaxies with @xmath90 . at the same time
, the systematically larger amplitude in the redder band indicates the presence of a wavelength - dependent effect which could be attributed to dust extinction . before interpreting these signals
, we note that both magnification and reddening effects are observed over a wide range of scales , from about 5to a few degrees .
we have therefore expanded the angular range used in the detection of magnification by @xcite to both smaller and larger separations . using the mean galaxy redshift computed in equation [ eq : mean_z ] an angular scale of one arcminute corresponds to roughly 150 @xmath91 kpc .
the above angular range thus corresponds to physical scales from @xmath92 @xmath91kpc to @xmath93 @xmath91mpc , as indicated in the top abscissa of the figure . in the inset of figure [ fig : mag_shift_log ]
, we show the wavelength dependence of the signal on small - scale with measurements of @xmath94 with the @xmath95 and @xmath96 filters .
we can observe a continuous trend as a function of wavelength : the signal is systematically stronger in redder bands , consistent with reddening by dust .
this also shows that probing magnification requires a correction for the effects of dust extinction .
based on the same quasar sample and a slightly brighter galaxy sample ( selected with @xmath97 ) @xcite used a density - based estimator and showed that the dominant magnification signal of the quasar - galaxy correlation follows the expected dependence as a function of quasar magnitude .
no such behavior could be detected when using stars instead of quasars as a control sample .
the amplitude of dust reddening depends only on the properties and amount of dust around the foreground galaxies selected for the cross - correlation .
we can recover a similar wavelength - dependent extinction as in figure [ fig : mag_shift_log ] or the mean reddening signal presented in figure [ fig : colors_line ] by using subsamples of quasars in different magnitude ranges . as a further sanity check
, we also replaced our quasars with stars selected to have the same magnitude and spatial distribution on the sky . measuring the cross - correlation between star brightness and galaxy overdensity with this sample
, we find no appreciable color excess for galaxies with @xmath98 and stars selected in various magnitude ranges . )
this effect is likely to be attributed to a small contamination of faint galaxies in the stellar sample at faint magnitudes where star - galaxy classification is incomplete .
this is expected to occur at @xmath99 .
the presence of galaxies in the star sample may give rise to a similar correlation between source reddening and galaxy overdensity , since galaxy clustering gives rise to an excess of redder galaxies in overdense regions , which produces a similar signal that mimics the reddening by dust .
we have measured the reddening - clustering correlation of sdss galaxies with @xmath100 and found that a contamination of galaxies at about 10% could explain the reddening signal seen around stars . ] finally , we have investigated the dependence of the signal as a function of galactic reddening : by splitting the dataset into two regimes of galactic extinction given by the @xcite map , we did not detect any significant change in our signal .
we isolate the reddening effects by measuring the correlation between qso color and foreground galaxy overdensity , @xmath101 , where @xmath82 and @xmath102 indicate two different pass bands .
we estimate the errors by computing the color covariance matrix from bootstrap resampling .
note that the errors on colors are smaller than the errors on brightness changes .
as figure [ fig : colors_line ] shows , quasar colors and galaxy overdensities are positively correlated , i.e. quasars appear to be redder when closer to high concentrations of foreground galaxies .
the reddening effects are detected for all color combinations , from @xmath103 to about @xmath104 , corresponding to physical scales ranging from 50 @xmath91kpc to about 10 @xmath91mpc .
the measurement probes galactic radii well beyond the size of galactic disks .
the amplitude of the effect is stronger in bluer bands .
we find that a background source whose light passes at around 20 @xmath91kpc from a foreground galaxy ( selected with @xmath90 ) will be , on average , redder by @xmath105 mag . .
note that @xmath106 is poorly constrained on small scales.,height=340 ] the five sdss filters allow us to constrain the shape of @xmath107 , the extinction curve of the dust associated with the galaxies , through four independent colors .
we measure the @xmath108 correlations for two angular bins : @xmath109 arcmin and @xmath110 arcmin , which correspond to effective projected radii of about @xmath111 @xmath91kpc and @xmath112mpc . in figure
[ fig : color_one_arcminute ] we show the corresponding color excesses with respect to the @xmath70 band .
we compare these reddening measurements to the standard extinction curve by fitting these data points with the functional form of the extinction curve provided by @xcite .
such extinction curves are usually characterized by the parameter @xmath113 , which characterizes the slope of the extinction curve .
the coefficient @xmath114 quantifies the amount of dust through its extinction in the @xmath115 band .
the best fit for @xmath114 and @xmath106 is shown with the blue curve . on small scales
our measured reddening corresponds to @xmath116 mag and @xmath117 , i.e. the slope of the extinction curve is not well constrained . on large scales
however we obtain a better accuracy : @xmath118 and @xmath119 mag .
the green dashed curve shows the best fit for @xmath114 when @xmath106 is assumed to be 3.1 , the standard value for dust in the disk of our galaxy . within the errors our results are consistent with standard interstellar dust .
given our measurements of the reddening due to foreground galaxies , we can now make estimates of the average dust extinction profile of the galaxies in our sample .
in addition to constraining the amount and spatial distribution of dust on large scales around galaxies , this quantity is of interest to quantify the intrinsic brightness of background sources , which is the main goal of type ia supernovae measurements aimed at constraining dark energy . as indicated in [ sec : extinction ] , the correlation between quasar color and galaxy overdensity is an estimator of the mean reddening induced by galaxies . in order to maximize the signal - to - noise ratio of the reddening detection
, we measure the quantity @xmath101 for which @xmath82 and @xmath102 are taken to be the @xmath47 and @xmath86 pass bands .
we avoid using the @xmath120 and @xmath96 bands for which the photometric errors are substantially larger .
in addition , the latter suffers from photometric contamination due to sky emission lines .
we show the mean @xmath121 color excess of quasars as a function of scale in figure [ fig : av ] .
a best fit power - law gives @xmath122 this quantity is proportional to the dust surface density and therefore provides us with direct constraints on the spherically average distribution of dust around galaxies .
the angular dependence is similar to average mass profiles around galaxies constrained from galaxy - galaxy lensing ( e.g. @xcite ) and magnification measurements by @xcite .
this result has a number of implications regarding the amount and nature of the dust in galaxy halos .
it will be discussed below in section [ sec : implications ] .
+ in order to convert reddening into extinction , we choose the value @xmath123 , corresponding to standard interstellar dust in our galactic disc and in agreement with the constraints obtained above . in figure
[ fig : av ] , we plot the mean extinction profile around the galaxies in our sample .
it can be written as @xmath124 or @xmath125 where @xmath114 is the _ observed _ @xmath115-band extinction .
for an extinction curve characterized by @xmath126 , we have @xmath127 in the visible range .
we point out that the amount of reddening profile shown in figure [ fig : av ] is significantly lower than contributions expected from dwarf galaxies .
for example , the average dust reddening induced by the lmc , located at about 50 kpc from our galaxy , is @xmath128 @xcite .
such a value is about an order of magnitude larger than eq .
[ eq : egi ] and shows that satellite galaxies are expected to dominate the amount of dust reddening for individual lines - of - sight intercepting galaxy halos .
the correlation between quasar magnitude shift and galaxy overdensity allows us to constrain a combination of magnification and dust extinction .
having estimated the wavelength - dependent extinction from reddening measurements in the previous section , we can now use this information and isolate the magnification effects .
our goal here is to estimate the signal well enough to allow for a comparison with similar measurements from galaxy - galaxy lensing available in the literature . for simplicity
we start by describing the angular correlation between quasar brightness and galaxy overdensity by a simple power - law with index @xmath129 . as above
, we use a value of @xmath123 ( or @xmath127 in the visible range ) to convert reddening into extinction . the mean quasar brightness change as
a function of angular separation from galaxies then reads @xmath130 \,\left(\frac{\theta}{1\arcmin}\right)^{-0.8}\ ; , \label{eq : fit_delta_m_obs}\ ] ] where @xmath131 is the achromatic magnitude change due to magnification and @xmath132 is the change due to dust extinction in the v - band . as shown above , we have @xmath133 ( eq . [ eq : av ] ) , which gives @xmath134 . at sufficiently large wavelength ,
dust extinction becomes negligible and the brightness change is only due to magnification . in this limit
, we find @xmath135 where we have used the conversion between observed and induced magnitude shift ( equation [ eq : get_mu ] ) and @xmath136 for sdss quasars with @xmath137 ( equation [ eq : cs ] ) .
this implies that the mean magnification excess found at one arcminute ( or @xmath0100 @xmath138kpc ) is about 2% .
this result is in agreement with that of @xcite where the authors inferred the magnification from the observed correlation between quasar and galaxy densities on the sky , i.e. a different estimator .
our analysis also allows us to measure magnification effects on smaller scales . on the smallest scale we can probe , i.e. 5(or @xmath015 @xmath138kpc ) , we find @xmath139 .
+ in the weak lensing regime , the magnification is related to the mean mass surface density by @xmath140 the inferred mean mass surface density of our galaxy sample is shown with the blue data points in figure [ fig : mu_gamma ] .
a best fit power - law distribution gives @xmath141 with @xmath142 , or using our effective projected scale : @xmath143 with @xmath144 .
these results provide an estimate of galaxy density profiles from gravitational magnification corrected for dust extinction effects .
while a linear relation between magnification and density contrast is a good approximation on large scales , higher - order corrections become significant on scales smaller than a few arcminutes . as shown by , using eq .
[ eq : magnification_linear ] results in overestimating the projected mass @xmath25 by about 15 to 25% on scales ranging from 1 to 0.1 arcminute .
the first few points shown in figure [ fig : mu_gamma ] have not been corrected for this effect . as shown below , on larger scales
the mass estimate from magnification is in good agreement with shear - based mass estimates .
so far the galaxy - mass cross - correlation has mostly been accessible through galaxy - galaxy lensing measurements which estimate the mean tangential shear of background galaxies for a given sample of foreground lenses ( e.g. @xcite , @xcite , @xcite , @xcite ) .
the tangential shear ( @xmath145 ) azimuthally averaged over a thin annulus at projected radius r from a lens galaxy is directly related to the projected surface mass density of the lens within the aperture , @xmath146 where @xmath147 @xmath148 is the mean surface density within radius @xmath149 , and @xmath150 is the azimuthally averaged surface density at radius @xmath149 ( @xcite ) .
therefore , shear measurements constrain @xmath151 whereas magnification is a direct estimate of @xmath25 . in order to compare magnification and shear measurements ,
it is interesting to note that , in the case of power - law mass profiles , with @xmath152 , we have @xmath153 the two observables @xmath25 and @xmath151 are therefore equivalent for isothermal profiles . for an angular dependence following @xmath154 , we have @xmath155 with @xmath156 .
this value goes down to @xmath157 for an index of @xmath158 . using galaxy - galaxy measurements based on the sdss
, @xcite found @xmath159 with @xmath160 and @xmath161 for a sample of spectroscopically identified lenses with @xmath162 and @xmath163 . considering for simplicity a slope of @xmath129 ( consistent with their constraint ) ,
their results translate into @xmath164 .
their scaled - measurements ( using equation [ eq : rescale_shear ] ) are shown in figure [ fig : mu_gamma ] . under these assumptions , the magnification and shear estimators appear to be in very good agreement .
a more accurate comparison between the two would require accounting for the differing luminosity distributions between the two lensing samples as well possible redshift evolution ( our lenses comprising most of the sources for the shear - based estimators ) .
however , it illustrates that both methods offer comparable degree of measurement precision and dynamic range while affected by significantly different systematic effects .
our detection of the change in apparent magnitude of distant quasars has allowed us to quantify the magnification and reddening of background sources as their light rays pass in the vicinity of foreground galaxies .
while the amplitude of gravitational lensing effects has been known both from theory and observations @xcite , the expected amplitude of dust reddening and extinction effects was largely unconstrained .
our study has shown that , in the visible bands , dust extinction occurs at a level comparable to that of the observed brightening due to magnification .
hence , studies aimed at predicting observable magnitude changes of background sources ( quasars , galaxies , supernovae , etc . ) which include only gravitational lensing effects are incomplete . in the visible range ,
dust extinction effects can not be neglected and must be included alongside with magnification to properly account for the effects of passing through large scale structure .
+ below , in order to simplify the discussion we will assume the dust in galactic halos to be described by smc type dust .
this choice is motivated by a number of results : ( i ) certain low - ionisation absorbers such as mgii are known to inhabit the halo of @xmath8 galaxies .
they are found on scales reaching up to about 100 @xmath138kpc around galaxies @xcite .
recently several authors studied their extinction properties and dust content @xcite .
they found that their average extinction curve is similar to that of the smc , i.e. does not show the 0.2 @xmath165 bump seen in the milky way extinction curve ( for systems with @xmath166 where the feature enters the visible window ) .
( ii ) in addition , it is known that only a small fraction of high redshift galaxies show an extinction curve with the @xmath167 bump . as discussed below , considering a milky - way type dust changes our results only by a factor two .
+ in order to characterize the population of galaxies responsible for most of the intergalactic dust , we first assume that the amount of dust in halos is , on average , proportional to the metallicity and luminosity of galaxies . under this assumption , the dominant contribution of the reddening signal is expected to originate from galaxies with an effective luminosity : @xmath168 where @xmath169 is the galaxy luminosity density given by the schechter function @xmath170 with @xmath171 and @xmath172 is the metallicity - luminosity relation @xcite : @xmath173 with @xmath174 . by integrating equation
[ eq : l_mean ] between @xmath175 and infinity , we find that the most important contribution of reddening originates from galaxies with a luminosity l_eff0.45 l^. [ eq : def_l ] given the number density of @xmath176 galaxies @xcite and the above luminosity function , galaxies with @xmath177 are expected to have a comoving number density of @xmath178 .
+ below we discuss a number of implications given by the existence of a large - scale distribution of dust around galaxies .
we note that only certain of the following results will make use of the above numbers . the properties of dust particles giving rise to the observed optical extinction and infrared emission in our galaxy , the lmc and the smc have been modeled by several authors ( e.g. @xcite and @xcite ) .
these models provide estimates of the absorption cross section per mass of dust as a function of wavelength @xmath179 , which is used to infer dust mass surface density from reddening profiles .
for the smc type dust the model of @xcite gives @xmath180 note that milky way type dust corresponds to a dust mass larger by a factor two at a fixed extinction .
the knowledge of @xmath181 allows us to convert the observed reddening into a dust mass surface density .
the spatial dependence of this quantity is shown in figure [ fig : mass_profiles ] .
we find a column density of dust of about @xmath182 at an impact parameter of 100 @xmath138kpc .
it is interesting to estimate the total amount of dust in the halo of the galaxies defined by equation [ eq : def_l ] . considering an isothermal sphere mass distribution ,
the virial radius of @xmath183 galaxies is @xmath184 .
the mass of dust residing in the halo of such galaxies is given by integrating the ratio @xmath185 over the area enclosed by @xmath186 : @xmath187 our measurements provide us with the radial dependence of @xmath114 for @xmath188 @xmath138kpc , i.e. scales greater than galactic disks . by defining the dust mass in the halo as @xmath189 and using eq .
[ eq : av ] with smc type dust , we find @xmath190 @xmath138kpc , this quantity is a direct estimate of the mean surface density of galaxies ( with @xmath90 ) . ]
this estimate shows that a substantial amount of dust exists in the halo of @xmath8 galaxies .
this dust mass is comparable to that commonly found in galactic disks @xcite .
this immediately shows that the reddening profile shown in figure [ fig : av ] can not be explained by a contribution from satellite galaxies whose luminosities and therefore dust masses is negligible compared to a central @xmath176 galaxy .
our results suggest the existence of a diffuse component of dust in galactic halos .
we note that similar arguments can be obtained from reddening considerations and do not rely on the specific value of @xmath181 ( eq . [ eq : k ] ) .
it should also be noted that while the total amount of dust appears to be dominated by a diffuse component , the largest reddening values are still expected to originate from lines of sight passing through satellites .
the average reddening value of the lmc , @xmath128 , is about one order of magnitude larger than the average halo reddening at a radius on order 50 @xmath138kpc ( see figure [ fig : av ] ) .
having obtained constraints on both the total mass distribution from magnification and dust mass from reddening ( see figure [ fig : mass_profiles ] ) , we can compare their statistical spatial distributions by computing the ratio @xmath191 where @xmath192 and @xmath25 are the dust mass and total mass surface densities .
this ratio is plotted as a function of scale in the lower panel of figure [ fig : mass_profiles ] and appears to be only weakly scale - dependent .
we find the dust to mass ratio @xmath193 for smc type dust .
we now attempt to estimate the cosmic density of dust , @xmath194 . to do so we first compute the density of dust originating from galaxy halos .
considering @xmath8 galaxies with a mass - to - light ratio @xmath195 @xcite , a light density @xmath196 @xcite and for a dust - to - mass ratio @xmath197 in halos , we can write @xmath198 previous attempts to obtain observational constraints on the cosmic density of dust have focused on light - weighted estimates , i.e. dust related to disks .
for example , using the attenuation - inclination relation for galaxy discs and their associated central bulges , @xcite quantified the mean attenuation of a large sample of galaxies and estimated @xmath199 .
similarly , theoretical estimates have focused on dust associated with cold gas .
for example , @xcite estimated the cosmic dust density by computing the mean metallicity of galaxies weighted by the schechter luminosity function , used a mass fraction of metals into dust grains of @xmath200 multiplied by the density parameter in cool gas .
they find @xmath201 .
the total cosmic dust density can be estimated by summing up the halo and disk contributions .
we find that @xmath202 this value is in agreement with the ( model - dependent ) upper limit obtained by @xcite : @xmath203 at @xmath204 .
it is about a factor two larger than the estimate given by @xcite for galactic disks . as our results provide us with an estimate of the mean optical depth for dust extinction due to a population of galaxies spanning a given redshift range , we can attempt to integrate @xmath205 over redshift and estimate the opacity of the universe .
extinction as a function of source redshift .
the blue circle shows our ( model - dependent ) estimate of the opacity induced by dust in the halo of @xmath206 galaxies at the mean redshift of our sample .
the three blue curves represent extrapolations to higher redshifts , using different dust models : a constant co - moving density ( dashed line ) , a density decreasing with the mean metallicity ( solid light blue ) and a suppressed density at high redshift ( dark blue ) .
various observational constraints are shown and described in appendix b. ] assuming that most of the dust in the universe is associated with galaxies and that most of the optical depth for extinction originates from the halo of galaxies with @xmath207 ( defined in equation [ eq : def_l ] ) , the mean dust optical depth in the universe up to a redshift @xmath96 is then given by @xmath208 where @xmath10 and @xmath209 are the cross - section and number density of galaxies with @xmath207 .
for a dust profile extending up to the virial radius , @xmath186 , of these galaxies we have @xmath210 and @xmath211 is the average optical depth within the halo , which we define with @xmath212 .
we evaluate @xmath213 ( equation [ eq : tau ] ) and present the results in terms of extinction in the observer - frame @xmath115-band . integrating equation [ eq : tau ] up to the mean redshift of the galaxy sample used in this analysis , we find the value denoted by the blue circle : @xmath214 .
we then estimate @xmath215 for different dust models shown in figure [ fig : cosmic_av ] : 1 .
we first use a constant dust density with redshift , represented with the dashed line . at high redshift , such an estimate is expected to represent an upper limit on the allowed extinction .
2 . a more realistic estimate should take into account the fact that the amount of dust in and around galaxies is redshift dependent as dust is being produced at a rate following that of the metals . as mentioned above , useful constraints on the evolution of the amount of dust around galaxies come from studies of metal absorbers .
recently , @xcite probed the redshift evolution of the amount of dust associated with mgii absorbers from @xmath216 to @xmath217 .
they found that their dust content follows @xmath218 , which turns out to be similar to the evolution of cosmic star density . by taking this redshift dependence into account
, we obtain a an alternative estimate of the cosmic opacity .
the corresponding results are shown in figure [ fig : cosmic_av ] with the solid light - blue curve .
as can be seen , taking this effect into account brings a modest change to our previous estimate and lowers the total opacity by about a factor two at @xmath219 .
3 . finally , to illustrate the range of possible redshift dependences , we consider a third model where we damp the density of dust by an addition factor @xmath220 , shown with the dark blue curve .
this model is somewhat _ ad hoc _ in nature , but demonstrates that the current measurements and limits allow for considerable variation at higher redshift .
figure [ fig : cosmic_av ] also shows various upper and lower limits ( detailled in appendix b ) on the opacity as a function of redshift .
as can be seen , at @xmath221 , @xmath114 values are bounded within about a factor @xmath222 .
type ia supernovae provide us with an estimate of luminosity distances and have been extensively used to constrain cosmological parameters , dark energy in particular .
supernovae are standardizable candles and become usable as standard candles after re - normalizing their brightness for intrinsic brighter - bluer and brighter - slower trends as well as dust extinction .
we now investigate how the presence of intergalactic dust can affect such constraints .
the color @xmath223 of each supernovae is the sum of several contributions : @xmath224 , where @xmath225 are the intrinsic color , dust reddening by the host , dust along the line - of - sight , etc .
each of them can be corrected for using an appropriate `` reddening - to - extinction '' coefficient @xmath226 .
the observed supernova magnitudes are used as a distance estimator according to @xmath227 where the apparent magnitude @xmath228 , the stretch @xmath229 and the color @xmath225 are derived from the fit to the light curves .
the parameters @xmath82 , @xmath102 and the absolute magnitude @xmath230 are fitted by minimizing the residuals in the hubble diagram . a color excess @xmath225 which does not contribute to a significant scatter
will not affect the inferred value of @xmath102 .
its reddening - to - extinction coefficient will be described by the best - fit @xmath231 and might lead to a bias if its intrinsic @xmath102 differs from this value .
as shown above , the presence of intergalactic dust might introduce such an effect . on large scales around galaxies
we find @xmath232 , which corresponds to a value of @xmath233 , while sn analyses lead to @xmath234 @xcite .
this indicates that , at some level , the ( redshift - dependent ) contribution of intergalactic might not be properly corrected by using eq .
[ eq : sn ] .
the magnitude of this effect and its impact on dark energy constraints must be investigated and quantified .
we can compare the statistical properties of the light ( traced by galaxies ) and dust distributions by computing the ratio of the galaxy - reddening cross - correlation to the autocorrelation of the same galaxies .
to do so we compute , as a function of scale , the parameter @xmath235 where we use the @xmath236 color as our estimate of dust reddening ( see section [ sec : profile ] ) .
this quantity is shown in figure [ fig : color_bias ] .
the dust reddening - to - light ratio appears to be weakly scale - dependent . on scales greater than about one arcminute
, we find @xmath237 the excess seen on the smallest scale might be due to a contribution from galactic disks .
the dust - to - light ratio is a quantity which can be used to constrain models describing the transport of dust outside of galaxies .
our results show that , on large scales , the dust distribution follows that of galaxies .
intervening galaxy halos and large - scale structure affect the light of background sources through both gravitational lensing and dust extinction .
in this paper we present simultaneous detections of these two effects obtained by measuring the cross - correlation between the brightness and colors of about 85,000 @xmath1 quasars and 20 million @xmath2 galaxies from the sdss .
we find that quasar brightness is correlated with galaxy overdensity , indicating that magnification effects dominate over dust extinction for this sample of galaxies . by examining the quasar colors ,
we observe that quasars appear to be redder as the projected galaxy density increases .
both effects are detected on scales ranging from 0.1 arcmin to about 2 degrees , corresponding to projected radii of 20@xmath138kpc to @xmath238 at the mean redshift of the galaxy sample .
more specifically , we find that * galaxies and large - scale structures at @xmath204 induce an excess reddening to background sources given by @xmath239 . on arcminute - scales around galaxies , the slope of the extinction curve is constrained by @xmath240 , a value consistent with that of our galaxy . assuming the galactic value of @xmath126 , we find a mean extinction profile given by @xmath241 .
the ratio of the galaxy - reddening to galaxy - galaxy correlation functions is found to be roughly scale independent , with an amplitude @xmath242 mag . *
the amplitude of dust extinction in the @xmath115-band is about one - third of that of the _ observed _ brightening due to magnification .
we estimate the extinction - corrected magnification profile and find @xmath243 , consistent with the results of @xcite and extending those to both smaller and larger scales .
the average mass surface density profile of the galaxies inferred from our measurements is comparable to galaxy - galaxy lensing estimates .
* at a separation of @xmath244 @xmath138kpc from a @xmath204 galaxy , a background source is , on average , magnified by a factor @xmath245 and reddened by e(b - v)@xmath246 .
the detection of dust on large - scales has a number of implications : * the amount of dust found in galactic halos is found to be comparable to that in the disk of @xmath8 galaxies and therefore significantly larger than that of dwarf satellites .
this implies the existence of a diffuse component of dust in halos , predicted by some models of dust halo dynamics but heretofore unobserved .
* having argued that the dominant contribution of dust reddening observed in this analysis is due to galaxies with @xmath247 , we have estimated their halo dust mass to be about @xmath248 within their virial radius . * including both disk and halo contributions , we find @xmath249 , a value roughly twice that estimated by @xcite for galactic disks .
* such an extended distribution of dust around galaxies will affect the apparent magnitude of distant sources .
we have estimated the mean opacity of the universe due to dust in galactic halos .
our model - dependent estimate gives @xmath250 mag .
such a value is less constrained at higher redshifts due to our limited knowledge of the evolution of dust with redshift .
considering several models , we found @xmath251 mag .
this will affect the brightness estimates of type ia supernovae at high redshift , which require high precision in order to maximize their constraints on cosmological parameters .
dust reddening may also induce dispersion in brightness and color .
the technique presented in this paper provides us with a unique probe of the distribution of dust ( warm and cold ) on large scales around galaxies , which is otherwise difficult to explore .
it opens up the way to studies of the amount of circumgalactic dust as a function of galaxy type , luminosity and environment which may shed light on the origin of the dust .
extending the analysis to uv measurements is an important task to obtain better constraints on the dust properties .
the final remark is that this analysis requires only accurate photometric data in several passbands .
we thank robert lupton , jim gunn , joseph weingartner , bruce draine , doron chelouche , tony tyson , erin sheldon , and latham boyle for useful discussions .
mf is supported by the monell fundation and the friends of the institute for advanced study .
funding for the sdss and sdss - ii has been provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , the u.s .
department of energy , the national aeronautics and space administration , the japanese monbukagakusho , the max planck society , and the higher education funding council for england .
the sdss web site is http://www.sdss.org/.
the observable magnitude shift defined by equation [ eq : obs ] is a function of the shape of the magnitude distribution , the limiting magnitude and the induced magnitude shift @xmath39 . in the case where the induced magnitude shift @xmath39 is small compared to the limiting magnitude of the sample , the difference between the observed and induced magnitude shift can be linearized in @xmath39 .
we have @xmath253 for induced magnitude shifts small compared to unity , we can taylor - expand the above expression to first - order in @xmath39 : @xmath254\,\end{aligned}\ ] ] where @xmath255 .
the above expression can be simply written as @xmath256 with @xmath42\ ] ] we can verify that in the case of a power - law luminosity function , i.e. @xmath257 or @xmath258 , the above expression gives @xmath259 which implies that no magnitude shift can be observed .
in this appendix , we detail the values of the cosmic opacity upper and lower limits used in figure [ fig : cosmic_av ] . note that the upper limit values plotted in the figure reflect the 95% or 99% confidence limits for those measurements , while the lower limit values are taken from the actual measured values .
* @xcite put a virtually assumption - free constraint on the opacity of the universe at low redshift using the tolman test : @xmath260 where @xmath261 is the luminosity distance and @xmath262 the angular diameter distance , independent of world model . any observed deviation from this expected relation
is taken to be a result of extinction along the line of sight .
combining observational results from supernovae and baryon acoustic oscillations to estimate the change in optical depth from redshift 0.20 to 0.35 , they found that @xmath263 at 95% confidence . assuming no evolution of the dust properties in the redshift range @xmath264
, we can use their result to estimate the expected @xmath114 extinction up to @xmath265 : @xmath266\end{aligned}\ ] ] where @xmath267 .
this gives @xmath268 mag .
* @xcite obtained a constraint on the low - redshift opacity by combining observational results from supernovae and the hubble key project .
they simultaneously fitted for @xmath269 and the luminosity distance , where the latter quantity was allowed to be modulated extinction . using the union sample of supernovae @xcite with a mean redshift of @xmath270
, they found @xmath271 with @xmath272 .
this gives @xmath273 mag . at 99% confidence .
* @xcite analyzed the scatter in quasar colors as a function of redshift , attributing its excess to dust extinction .
they reported a @xmath274 upper limit on the cosmic opacity : @xmath275 mag .
* @xcite quantified the mean amount of reddening and extinction induced by strong mgii absorbers , i.e. systems usually found within @xmath276 kpc of @xmath8 galaxies @xcite .
they were able to constrain the mean reddening @xmath277 as a function of the rest equivalent width @xmath278 of the absorbers and their redshift , in the range @xmath279 .
in addition , as described in section [ sec : implications ] , various authors have shown that , on average , the extinction curve associated with mgii absorbers is consistent with that of the smc , i.e. with @xmath280 .
the mean extinction induced by these systems can then be computed according to @xmath281 using the parameterization of the the incidence rate of mgii absorbers , @xmath282 , given by @xcite , and assuming no evolution in the dust properties in the range @xmath283 , we can use these measured values as lower limits on the global extinction .
we find : @xmath284 mag , @xmath285 mag and @xmath286 mag . | we present a simultaneous detection of gravitational magnification and dust reddening effects due to galactic halos and large - scale structure .
the measurement is based on correlating the brightness of @xmath085,000 quasars at @xmath1 with the position of 20 million galaxies at @xmath2 derived from the sloan digital sky survey and is used to constrain the galaxy - mass and galaxy - dust correlation functions up to cosmological scales .
the presence of dust is detected from 20 kpc to several mpc , and we find its projected density to follow : @xmath3 , a distribution similar to mass .
the amount of dust in galactic halos is found to be comparable to that in disks . on large scales
its wavelength dependence is described by @xmath4 , consistent with interstellar dust .
we estimate the resulting opacity of the universe as a function of redshift and find @xmath5 mag up to @xmath6 .
this , in turn , implies a cosmic dust density of @xmath7 , roughly half of which comes from dust in halos of @xmath8 galaxies .
we present magnification measurements , corrected for dust extinction , from which the galaxy - mass correlation function is inferred . the mean mass profile around galaxies is found to be @xmath9 up to a radius of 10 mpc , in agreement with gravitational shear estimates .
dust extinction , reddening
dark matter magnification large - scale structures quasars galaxies |
bandwidth selection is a key issue in kernel density estimation that has deserved considerable attention during the last decades .
the problem of selecting the most suitable bandwidth for the nonparametric kernel density estimator introduced by @xcite and @xcite is the main topic of the reviews of @xcite , @xcite and @xcite , among others .
comprehensive references on kernel smoothing and bandwidth selection include the books by @xcite , @xcite and @xcite .
bandwidth selection is still an active research field in density estimation , with some recent contributions like @xcite and @xcite in the last years . + kernel density estimation has been also adapted to directional data , that is , data in the unit hypersphere of dimension @xmath0 .
due to the particular nature of directional data ( periodicity for @xmath1 and manifold structure for any @xmath0 ) , the usual multivariate techniques are not appropriate and specific methodology that accounts for their characteristics has to be considered .
the classical references for the theory of directional statistics are the complete review of @xcite and the book by @xcite .
the kernel density estimation with directional data was firstly proposed by @xcite , studying the properties of two types of kernel density estimators and providing cross - validatory bandwidth selectors .
almost simultaneously , @xcite provided a similar definition of kernel estimator , establishing its pointwise and @xmath2 consistency .
some of the results by @xcite were extended by @xcite , who studied the estimation of the laplacian of the density and other types of derivatives . whereas the framework for all these references is the general @xmath0-sphere , which comprises as particular case the circle ( @xmath1 ) , there exists a remarkable collection of works devoted to kernel density estimation and bandwidth selection for the circular scenario .
specifically , @xcite presented the first plug - in bandwidth selector in this context and @xcite derived a selector based on mixtures and on the results of @xcite for the circular asymptotic mean integrated squared error ( amise ) .
recently , @xcite proposed a product kernel density estimator on the @xmath0-dimensional torus and cross - validatory bandwidth selection methods for that situation .
another nonparametric approximation for density estimation with circular data was given in @xcite and @xcite .
in the general setting of spherical random fields @xcite derived an estimation method based on a needlet basis representation .
+ directional data arise in many applied fields . for the circular case ( @xmath1 ) a typical example is wind direction , studied among others in @xcite , @xcite and @xcite .
the spherical case ( @xmath3 ) poses challenging applications in astronomy , for example in the study of stars position in the celestial sphere or in the study of the cosmic microwave background radiation @xcite . finally , a novel field where directional data is present for large @xmath0 is text mining @xcite , where documents are usually codified as high dimensional unit vectors .
for all these situations , a reliable method for choosing the bandwidth parameter seems necessary to trust the density estimate .
+ the aim of this work is to introduce new bandwidth selectors for the kernel density estimator for directional data .
the first one is a rule of thumb which assumes that the underlying density is a von mises and it is intended to be the directional analogue of the rule of thumb proposed by @xcite for data in the real line .
this selector uses the amise expression that can be seen , among others , in @xcite .
the novelty of the selector is that it is more general and robust than the previous proposal by @xcite , although both rules exhibit an unsatisfactory behaviour when the reference density spreads off from the von mises . to overcome this problem , two new selectors based on the use of mixtures of von mises for the reference density
are proposed .
one of them uses the aforementioned amise expression , whereas the other one uses the exact mise computation for mixtures of von mises densities given in @xcite .
both of them use the expectation - maximization algorithm of @xcite to fit the mixtures and , to select the number of components , the bic criteria is employed .
these selectors based on mixtures are inspired by the earlier ideas of @xcite , for the multivariate setting , and @xcite for the circular scenario .
+ this paper is organized as follows .
section [ kdebwd : sec : kdedir ] presents some background on kernel density estimation for directional data and the available bandwidth selectors .
the rule of thumb selector is introduced in section [ kdebwd : sec : ruleofthumb ] and the two selectors based on mixtures of von mises are presented in section [ kdebwd : sec : mixtures ] .
section [ kdebwd : sec : comparative ] contains a simulation study comparing the proposed selectors with the ones available in the literature .
finally , section [ kdebwd : sec : data ] illustrates a real data application and some conclusions are given in section [ kdebwd : sec : conclusions ] .
supplementary materials with proofs , simulated models and extended tables are given in the appendix .
denote by @xmath4 a directional random variable with density @xmath5 .
the support of such variable is the @xmath0-dimensional sphere , namely @xmath6 , endowed with the lebesgue measure in @xmath7 , that will be denoted by @xmath8 .
then , a directional density is a nonnegative function that satisfies @xmath9 .
also , when there is no possible confusion , the area of @xmath7 will be denoted by @xmath10 where @xmath11 represents the gamma function defined as @xmath12 , @xmath13 .
+ among the directional distributions , the von mises - fisher distribution ( see @xcite ) is perhaps the most widely used .
the von mises density , denoted by @xmath14 , is given by @xmath15 where @xmath16 is the directional mean , @xmath17 the concentration parameter around the mean , @xmath18 stands for the transpose operator and @xmath19 is the modified bessel function of order @xmath20 , @xmath21 this distribution is the main reference for directional models and , in that sense , plays the role of the normal distribution for directional data ( is also a multivariate normal @xmath22 conditioned on @xmath7 ; see @xcite ) . a particular case of this density sets @xmath23 , which corresponds to the uniform density that assigns probability @xmath24 to any direction in @xmath7 .
+ given a random sample @xmath25 from the directional random variable @xmath4 , the proposal of @xcite for the directional kernel density estimator at a point @xmath26 is @xmath27 where @xmath28 is a directional kernel ( a rapidly decaying function with nonnegative values and defined in @xmath29 ) , @xmath30 is the bandwidth parameter and @xmath31 is a normalizing constant .
this constant is needed in order to ensure that the estimator is indeed a density and satisfies that @xmath32 as usual in kernel smoothing , the selection of the bandwidth is a crucial step that affects notably the final estimation : large values of @xmath33 result in a uniform density in the sphere , whereas small values of @xmath33 provide an undersmoothed estimator with high concentrations around the sample observations . on the other hand ,
the choice of the kernel is not seen as important for practical purposes and the most common choice is the so called von mises kernel @xmath34 . its name is due to the fact that the kernel estimator can be viewed as a mixture of von mises - fisher densities as follows : @xmath35 where , for each von mises component , the mean value is the @xmath36-th observation @xmath37 and the common concentration parameter is given by @xmath38 .
+ the classical error measurement in kernel density estimation is the @xmath39 distance between the estimator @xmath40 and the target density @xmath5 , the so called integrated squared error ( ise ) . as this is a random quantity depending on the sample , its expected value , the mean integrated squared error ( mise ) ,
is usually considered : @xmath41{\right]}}={\mathbb{e}{\left[}{\int_{\omega_{q } } { \left(\hat f_h({\mathbf{x}})-f({\mathbf{x}})\right)}^2\,\omega_{q}(d { \mathbf{x}})}{\right]}},\end{aligned}\ ] ] which depends on the bandwidth @xmath33 , the kernel @xmath28 , the sample size @xmath42 and the target density @xmath5 . whereas the two last elements are fixed when estimating a density from a random sample , the bandwidth has to be chosen ( also the kernel , although this does not present a big impact in the performance of the estimator ) .
then , a possibility is to search for the bandwidth that minimizes the mise : @xmath43 to derive an easier form for the mise that allows to obtain @xmath44 , the following conditions on the elements of the estimator ( [ kdebwd : kernel_directional ] ) are required : 1 .
extend @xmath5 from @xmath45 to @xmath46 by @xmath47 for all @xmath48 , where @xmath49 denotes the euclidean norm .
assume that the gradient vector @xmath50 and the hessian matrix @xmath51 exist , are continuous and square integrable .
[ kdebwd : cond : d1 ] 2 .
assume that @xmath52 is a bounded and integrable function such that @xmath53 , @xmath54 , for @xmath55.[kdebwd : cond : d2 ] 3 .
assume that @xmath56 is a positive sequence such that @xmath57 and @xmath58 as @xmath59.[kdebwd : cond : d3 ] the following result , available from @xcite , provides the mise expansion for the estimator ( [ kdebwd : kernel_directional ] ) .
it is worth mentioning that , under similar conditions , @xcite and @xcite also derived analogous expressions .
[ kdebwd : dir : prop:3 ] under conditions [ kdebwd : cond : d1][kdebwd : cond : d3 ] , the mise for the directional kernel density estimator ( [ kdebwd : kernel_directional ] ) is given by @xmath60 where @xmath61 , @xmath62 , @xmath63 and @xmath64 this results leads to the decomposition @xmath65 , where amise stands for the asymptotic mise .
it is possible to derive an optimal bandwidth for the amise in this sense , @xmath66 , that will be close to @xmath44 when @xmath67 is small enough .
[ kdebwd : dir : cor:1 ] the amise optimal bandwidth for the directional kernel density estimator ( [ kdebwd : kernel_directional ] ) is given by @xmath68}^{\frac{1}{4+q}},\label{kdebwd : dir : cor:1:1}\end{aligned}\ ] ] where @xmath69 .
unfortunately , expression ( [ kdebwd : dir : cor:1:1 ] ) can not be used in practise since it depends on the curvature term @xmath70 of the unknown density @xmath5 .
the first proposals for data - driven bandwidth selection with directional data are from @xcite , who provide cross - validatory selectors .
specifically , least squares cross - validation ( lscv ) and likelihood cross - validation ( lcv ) selectors are introduced , arising as the minimizers of the cross - validated estimates of the squared error loss and the kullback - leibler loss , respectively .
the selectors have the following expressions : @xmath71 where @xmath72 represents the kernel estimator computed without the @xmath36-th observation .
see remark [ kdebwd : rem:3 ] for an efficient computation of @xmath73 .
+ recently , @xcite proposed a plug - in selector for the case of circular data ( @xmath1 ) for the estimator with the von mises kernel .
the selector of @xcite uses from the beginning the assumption that the reference density is a von mises to construct the amise .
this contrasts with the classic rule of thumb selector of @xcite , which supposes at the end ( _ i.e. _ , after deriving the amise expression ) that the reference density is a normal .
the bandwidth parameter is chosen by first obtaining an estimation @xmath74 of the concentration parameter @xmath75 in the reference density ( for example , by maximum likelihood ) and using the formula @xmath76}^\frac{1}{5}.\end{aligned}\ ] ] note that the parametrization of @xcite has been adapted to the context of the estimator ( [ kdebwd : kernel_directional ] ) by denoting by @xmath33 the inverse of the squared concentration parameter employed in his paper . + more recently , @xcite proposed a selector that improves the performance of @xcite allowing for more flexibility in the reference density , considering a mixture of von mises .
this selector is also devoted to the circular case and is mainly based on two elements .
first , the amise expansion that @xcite derived for the circular kernel density estimator by the use of fourier expansions of the circular kernels .
this expression has the following form when the kernel is a circular von mises ( the estimator is equivalent to consider @xmath34 , @xmath1 and @xmath33 as the inverse of the squared concentration parameter in ( [ kdebwd : kernel_directional ] ) ) : @xmath77}^2\int_0^{2\pi } f''(\theta)^2\,d\theta+\frac{{\mathcal{i}}_0\big(2h^{-1/2}\big)}{2n\pi{\mathcal{i}}_0{\left(h^{-1/2}\right)}^2}. \label{kdebwd : dimarzio}\end{aligned}\ ] ] the second element is the expectation - maximization ( em ) algorithm of @xcite for fitting mixtures of directional von mises .
the selector , that will be denoted by @xmath78 , proceeds as follows :
1 . use the em algorithm to fit mixtures from a determined range of components .
choose the fitted mixture with minimum aic .
3 . compute the curvature term in ( [ kdebwd : dimarzio ] ) using the fitted mixture and seek for the @xmath33 that minimizes this expression , that will be @xmath78 .
using the properties of the von mises density it is possible to derive a directional analogue to the rule of thumb of @xcite , which is the optimal amise bandwidth for normal reference density and normal kernel .
the rule is resumed in the following result .
[ kdebwd : prop : rot ] the curvature term for a von mises density @xmath14 is @xmath79}.\end{aligned}\ ] ] if @xmath74 is a suitable estimator for @xmath75 , then the rule of thumb selector for the kernel estimator ( [ kdebwd : kernel_directional ] ) with a directional kernel @xmath28 is @xmath80}^{\frac{1}{4+q}}.\end{aligned}\ ] ] if @xmath28 is the von mises kernel , then : @xmath81}n}\right]}^\frac{1}{5 } , & q=1,\\ \displaystyle{\left[\frac{8\sinh^2(\hat\kappa)}{\hat\kappa{\left[(1 + 4\hat\kappa^2)\sinh(2\hat\kappa)-2\hat\kappa\cosh(2\hat\kappa)\right]}n}\right]}^\frac{1}{6 } , & q=2,\\ \displaystyle{\left[\frac{4\pi^\frac{1}{2}{\mathcal{i}}_{\frac{q-1}{2}}(\hat\kappa)^2}{\hat\kappa^{\frac{q+1}{2}}{\left[2q{\mathcal{i}}_{\frac{q+1}{2}}(2\hat\kappa)+(2+q)\hat\kappa{\mathcal{i}}_{\frac{q+3}{2}}(2\hat\kappa)\right]}n}\right]}^\frac{1}{4+q } , & q\geq3 .
\end{array}{\right.}\label{kdebwd : rot}\end{aligned}\ ] ] the parameter @xmath75 can be estimated by maximum likelihood . in view of the expression for @xmath82 in ( [ kdebwd : rot ] ) ,
it is interesting to compare it with @xmath83 when @xmath1 .
as it can be seen , both selectors coincide except for one difference : the term @xmath84 in the sum in the denominator of @xmath82 .
this `` extra term '' can be explained by examining the way that both selectors are derived . whereas the selector @xmath82 derives the bandwidth supposing that the reference density is a von mises when the amise is already derived in a general way , the selector @xmath83 uses the von mises assumption to compute it .
therefore , it is expected that the selector @xmath85 will be more robust against deviations from the von mises density .
+ figure [ kdebwd : fig : vs ] collects two graphs exposing these comments , that are also corroborated in section [ kdebwd : sec : comparative ] .
the left plot shows the mise for @xmath86 and @xmath82 for the density @xmath87 , where @xmath88 $ ] .
this model represents two equally concentrated von mises densities that spread off from being the same to being antipodal . as it can be seen , the @xmath85 selector is slightly more accurate when the von mises model holds ( @xmath89 ) and when the deviation is large ( @xmath90 $ ] ) .
when @xmath91 $ ] , both selectors perform similar .
this graph also illustrates the main problem of these selectors : the von mises density is not flexible enough to capture densities with multimodality and it approximates them by the flat uniform density . + when the density is a @xmath14 , the right plot of figure [ kdebwd : fig : vs ] shows the output of @xmath83 , @xmath85 , @xmath92 and their corresponding errors with respect to @xmath75 .
the effect of the `` extra term '' is visible for low values of @xmath75 , where @xmath93 presents a local maxima .
this corresponds with higher values of @xmath83 with respect to @xmath85 and @xmath92 , which means that the former produces oversmoothed estimations of the density ( _ i.e. _ tend to the uniform case faster ) . despite the worse behaviour of @xmath83 ,
when the concentration parameter increases the effect of the `` extra term '' is mitigated and both selectors are almost the same . .
left plot : logarithm of the curves of @xmath93 , @xmath94 and @xmath95 for sample size @xmath96 .
the curves are computed by @xmath97 monte carlo samples and @xmath92 is obtained exactly .
the abscissae axis represents the variation of the parameter @xmath98}$ ] , which indexes the reference density @xmath87 .
right plot : logarithm of @xmath83 , @xmath85 , @xmath92 and their corresponding mise for different values of @xmath75 , with @xmath96.[kdebwd : fig : vs],title="fig : " ] . left plot : logarithm of the curves of @xmath93 , @xmath94 and @xmath95 for sample size @xmath96 .
the curves are computed by @xmath97 monte carlo samples and @xmath92 is obtained exactly .
the abscissae axis represents the variation of the parameter @xmath98}$ ] , which indexes the reference density @xmath87 .
right plot : logarithm of @xmath83 , @xmath85 , @xmath92 and their corresponding mise for different values of @xmath75 , with @xmath96.[kdebwd : fig : vs],title="fig : " ]
the results of the previous section show that , although the rule of thumb presents a significant improvement with respect to the @xcite selector in terms of generality and robustness , it also shares the same drawbacks when the underlying density is not the von mises model ( see figure [ kdebwd : fig : vs ] ) . to overcome these problems , two alternatives for improving @xmath85 will be considered .
+ the first one is related with improving the reference density to plug - in into the curvature term .
the von mises density has been proved to be not flexible enough to estimate properly the curvature term in ( [ kdebwd : dir : cor:1:1 ] ) .
this is specially visible when the underlying model is a mixture of antipodal von mises , but the estimated curvature term is close to zero ( the curvature of a uniform density ) . a modification in this direction is to consider a suitable mixture of von mises for the reference density , that will be able to capture the curvature of rather complex underlying densities .
this idea was employed first by @xcite considering mixtures of multivariate normals and by @xcite in the circular setting .
+ the second improvement is concerned with the error criterion for the choice of the bandwidth . until now , the error criterion considered was the amise , which is the usual in the literature of kernel smoothing . however , as @xcite showed for the linear case and @xcite did for the directional situation , the amise and mise may differ significantly for moderate and even large sample sizes , with a potential significative misfit between @xmath99 and @xmath92 . then
, a substantial decreasing of the error of the estimator ( [ kdebwd : kernel_directional ] ) is likely to happen if the bandwidth is obtained from the exact mise , instead of the asymptotic version .
obviously , the problem of this new approach is how to compute exactly the mise , but this can be done if the reference density is a mixture of von mises .
+ the previous two considerations , improve the reference density and the error criterion , will lead to the bandwidth selectors of asymptotic mixtures ( ami ) , denoted by @xmath100 , and exact mixtures ( emi ) , denoted by @xmath101 . before explaining in detail the two proposed selectors , it is required to introduce some notation on mixtures of von mises .
+ an @xmath102-mixture of von mises densities with means @xmath103 , concentration parameters @xmath104 and weights @xmath105,with @xmath106 , is denoted by @xmath107 when dealing with mixtures , the tuning parameter is the number of components , @xmath102 , which can be estimated from the sample .
the notation @xmath108 will be employed to represent the mixture of @xmath109 components where the parameters are estimated and @xmath109 is obtained from the sample .
the details of this fitting are explained later in algorithm [ kdebwd : algo : nm ] .
+ then , the ami selector follows from modifying the rule of thumb selector to allow fitted mixtures of von mises .
it is stated in the next procedure .
[ kdebwd : algo : ami ] let @xmath25 be a random sample of a directional variable@xmath4 . 1 .
compute a suitable estimation @xmath108 using algorithm [ kdebwd : algo : nm ] .
2 . for a directional kernel @xmath28 ,
set @xmath110}^{\frac{1}{4+q}}\end{aligned}\ ] ] and for the von mises kernel , @xmath111}^{-\frac{1}{4+q}}.\end{aligned}\ ] ] [ kdebwd : rem:1 ] unfortunately , the curvature term @xmath112 does not admit a simple closed expression , unless for the case where @xmath113 , _
i.e. _ , when @xmath100 is equivalent to @xmath85 .
this is due to the cross - product terms between the derivatives of the mixtures that appear in the integrand .
however , this issue can be bypassed by using either numerical integration in @xmath0-spherical coordinates or monte carlo integration to compute @xmath112 for any @xmath114 .
the emi selector relies on the exact expression of the mise for densities of the type ( [ kdebwd : mise : mvm ] ) , that will be denoted by @xmath115}}.\end{aligned}\ ] ] similarly to what @xcite did for the linear case , @xcite derived the closed expression of @xmath116 when the directional kernel is the von mises one . the calculations are based on the convolution properties of the von mises , which unfortunately are not so straightforward as the ones for the normal , resulting in more complex expressions .
[ kdebwd : mise : th:1 ] let @xmath117 be the density of an @xmath102-mixture of directional von mises ( [ kdebwd : mise : mvm ] ) .
the exact mise of the directional kernel estimator ( [ kdebwd : kernel_directional ] ) with von mises kernel and obtained from a random sample of size @xmath42 is @xmath118}\mathbf{p } , \label{kdebwd : mise : mvm:1}\end{aligned}\ ] ] where @xmath119 and @xmath120 . the matrices @xmath121 , @xmath122 have entries : @xmath123 where @xmath124 is defined in equation ( [ kdebwd : dir : cq ] ) .
[ kdebwd : rem:2 ] a more efficient way to implement ( [ kdebwd : mise : mvm:1 ] ) , specially for large sample sizes and higher dimensions , is the following expression : @xmath125}}-f_m({\mathbf{x}})\right)}^2-{\mathbb{e}{\left[}\hat f_h({\mathbf{x}}){\right]}}^2\right\}}\,\omega_{q}(d { \mathbf{x}})},\end{aligned}\ ] ] where the integral is either evaluated numerically using @xmath0-spherical coordinates or monte carlo integration and @xmath126}}$ ] is computed using @xmath127}}&=\sum_{j=1}^m p_j \frac{c_q(\kappa_j)c_q{\left(1/h^2\right)}}{c_q\big(||{\mathbf{x}}/h^2+\kappa_j{\boldsymbol\mu}_j||\big)}.\end{aligned}\ ] ] [ kdebwd : rem:3 ] by the use of similar techniques , when the kernel is von mises , the lscv selector admits an easier expression for the cv@xmath128 loss that avoids the calculation of the integral of @xmath129 : @xmath130}-\frac{c_q(1/h^2)^2}{nc_q(2/h^2)}.\end{aligned}\ ] ] based on the previous result , the philosophy of the emi selector is the following : using a suitable pilot parametric estimation of the unknown density ( given by algorithm [ kdebwd : algo : nm ] ) , build the exact mise and obtain the bandwidth that minimizes it .
this is summarized in the following procedure .
[ kdebwd : algo : emi ] consider the von mises kernel and let @xmath25 be a random sample of a directional variable @xmath4 . 1 .
compute a suitable estimation @xmath108 using algorithm [ kdebwd : algo : nm ] .
2 . obtain @xmath131 .
the em algorithm of @xcite , implemented in the ` r ` package ` movmf ` ( see @xcite ) , provides a complete solution to the problem of estimation of the parameters in a mixture of directional von mises of dimension @xmath0 .
however , the issue of selecting the number of components of the mixture in an automatic and optimal way is still an open problem .
+ the propose considered in this work is an heuristic approach based on the bayesian information criterion ( bic ) , defined as @xmath132 , where @xmath133 is the log - likelihood of the model and @xmath134 is the number of parameters .
the procedure looks for the fitted mixture with a number of components @xmath102 that minimizes the bic .
this problem can be summarized as the global minimization of a function ( bic ) defined on the naturals ( number of components ) .
+ the heuristic procedure starts by fitting mixtures from @xmath135 to @xmath136 , computing their bic and providing @xmath114 , the number of components with minimum bic .
then , in order to ensure that @xmath114 is a global minimum and not a local one , @xmath137 neighbours next to @xmath114 are explored ( _ i.e. _ fit mixture , compute bic and update @xmath114 ) , if they were not previously explored .
this procedure continues until @xmath114 has at least @xmath137 neighbours at each side with larger bics .
a reasonable compromise for @xmath138 and @xmath137 , checked by simulations , is to set @xmath139 and @xmath140 . in order to avoid spurious solutions , fitted mixtures with any @xmath141
are removed .
the procedure is detailed as follows .
[ kdebwd : algo : nm ] let @xmath25 be a random sample of a directional variable @xmath4 with density @xmath5 . 1
. set @xmath139 and @xmath137 as the user supplies , usually @xmath140 .
[ kdebwd : algo : nm:1 ] 2 . for @xmath102 varying from @xmath142 to @xmath138 , [ kdebwd : algo : nm:2 ] 1 .
estimate the @xmath102-mixture with the em algorithm of @xcite and 2 .
compute the bic of the fitted mixture .
3 . set @xmath114 as the number of components of the mixture with lower bic .
[ kdebwd : algo : nm:3 ] 4 .
if @xmath143 , set @xmath144 and turn back to step [ kdebwd : algo : nm:2 ] .
otherwise , end with the final estimation @xmath108 .
[ kdebwd : algo : nm:4 ] other informative criteria , such as the akaike information criterion ( aic ) and its corrected version , aicc , were checked in the simulation study together with bic .
the bic turned out to be the best choice to use with the ami and emi selectors , as it yielded the minimum errors .
along this section , the three new bandwidth selectors will be compared with the already proposed selectors described in subsection [ kdebwd : subsec : bwsels ] . a collection of directional models , with their corresponding simulation schemes ,
are considered .
subsection [ kdebwd : subsec : dirmods ] is devoted to comment the directional models used in the simulation study ( all of them are defined for any arbitrary dimension @xmath0 , not just for the circular or spherical case ) .
these models are also described in the appendix . + for each of the different combinations of dimension , sample size and model , the mise of each selector was estimated empirically by @xmath97 monte carlo samples , with the same seed for the different selectors .
this is used in the computation of @xmath95 , where @xmath92 is obtained as a numerical minimization of the estimated mise .
the calculus of the ise was done by : simpson quadrature rule with @xmath145 discretization points for @xmath1 ; @xcite rule with @xmath146 nodes for @xmath3 and monte carlo integration with @xmath147 sampling points for @xmath148 ( same seed for all the integrations ) . finally , the kernel considered in the study is the von mises .
the first models considered are the uniform density in @xmath7 and the von mises density given in ( [ kdebwd : dir : cq ] ) .
the analogous of the von mises for axial data ( _ i.e. _ , directional data where @xmath149 ) is the watson distribution @xmath150 @xcite : @xmath151 where @xmath152 .
this density has two antipodal modes : @xmath153 and @xmath154 , both of them with concentration parameter @xmath17 .
a further extension of this density is the called small circle distribution @xmath155 @xcite : @xmath156 where @xmath157 , @xmath158 and @xmath159 . for the case
@xmath160 , this density has a kind of modal strip along the @xmath161-sphere @xmath162 . +
a common feature of all these densities is that they are rotationally symmetric , that is , their contourlines are @xmath161-spheres orthogonal to a particular direction .
this characteristic can be exploited by means of the so called tangent - normal decomposition ( see @xcite ) , that leads to the change of variables @xmath163 where @xmath164 is a fixed vector , @xmath165 ( measures the distance of @xmath166 from @xmath153 ) , @xmath167 and @xmath168 is the semi - orthonormal matrix ( @xmath169 and @xmath170 , with @xmath171 the @xmath0-identity matrix ) resulting from the completion of @xmath153 to the orthonormal basis @xmath172 .
the family of rotationally symmetric densities can be parametrized as @xmath173 where @xmath174 is a function depending on a vector parameter @xmath175 and such that @xmath176 is a density in @xmath177 , for all @xmath178 .
using this property , it is easy to simulate from ( [ kdebwd : rotsym ] ) . 1 .
sample @xmath179 from the density @xmath176.[kdebwd : algo : rotsym:1 ] 2 .
sample @xmath180 from a uniform in @xmath181 @xmath182.[kdebwd : algo : rotsym:2 ] 3 .
@xmath183 is a sample from @xmath184.[kdebwd : algo : rotsym:3 ] step [ kdebwd : algo : rotsym:1 ] can always be performed using the inversion method @xcite .
this approach can be computationally expensive : it involves solving the root of the distribution function , which is computed from an integral evaluated numerically if no closed expression is available .
a reasonable solution to this ( for a fixed choice of @xmath174 and @xmath153 ) is to evaluate once the quantile function in a dense grid ( for example , @xmath145 points equispaced in @xmath185 ) , save the grid and use it to interpolate using cubic splines the new evaluations , which is computationally fast . extending these ideas for rotationally symmetric models ,
two new directional densities are proposed .
the first one is the directional cauchy density @xmath186 , defined as an analogy with the usual cauchy distribution as @xmath187 where @xmath153 is the mode direction and @xmath17 the concentration parameter around it ( @xmath23 gives the uniform density ) .
this density shares also some of the characteristics of the usual cauchy distribution : high concentration around a peaked mode and a power decay of the density .
the other proposed density is the skew normal directional density @xmath188 , @xmath189 where @xmath190 is the skew normal density of @xcite with location @xmath191 , scale @xmath192 and shape @xmath193 that is truncated to the interval @xmath177 .
the density is inspired by the wrapped skew normal distribution of @xcite , although it is based on the rotationally symmetry rather than in wrapping techniques .
a particular form of this density is an homogeneous `` cap '' in a neighbourhood of @xmath153 that decreases very fast outside of it .
+ non rotationally symmetric densities can be created by mixtures of rotationally symmetric .
however , it is interesting to introduce a purely non rotationally symmetric density : the projected normal distribution of @xcite . denoted by @xmath194 , the corresponding densityis @xmath195 where @xmath196 , @xmath197 , @xmath198 and @xmath199 .
sampling from this distribution is extremely easy : just sample @xmath200 and then project @xmath4 to @xmath7 by @xmath201 .
+ the whole collection of models , with @xmath202 densities in total , are detailed in table [ kdebwd : tab : models ] in appendix [ kdebwd : ap : models ] .
figures [ kdebwd : fig : circ ] and [ kdebwd : fig : sph ] show the plots of these densities for the circular and spherical cases . for the circular case , the comparative study has been done for the @xmath202 models described in figure [ kdebwd : fig : circ ] ( see table [ kdebwd : tab : models ] to see their densities ) , for the circular selectors @xmath203 , @xmath204 , @xmath86 , @xmath78 , @xmath82 , @xmath205 and @xmath206 and for the sample sizes @xmath207 , @xmath208 , @xmath209 and @xmath97 . due to space limitations ,
only the results for sample size @xmath209 are shown in table [ kdebwd : tab : cir ] , and the rest of them are relegated to appendix [ kdebwd : ap : tables ] .
+ in addition , to help summarizing the results a ranking similar to ranking b of @xcite will be constructed .
the ranking will be computed according to the following criteria : for each model , the @xmath191 bandwidth selectors @xmath210 considered are sorted from the best performance ( lowest error ) to the worst performance ( largest error ) .
the best bandwidth receives @xmath191 points , the second @xmath211 and so on .
these points , denoted by @xmath212 , are standardized by @xmath191 and multiplied by the relative performance of each selector compared with the best one .
in other words , the points of the selector @xmath213 , if @xmath214 is the best one , are @xmath215 .
the final score for each selector is the sum of the ranks obtained in all the twenty models ( thus , a selector which is the best in all models will have @xmath202 points ) . with this ranking , it is easy to group the results in a single and easy to read table .
+ in view of the results , the following conclusions can be extracted .
firstly , @xmath85 performs well in certain unimodal models such as m3 ( von mises ) and m6 ( skew normal directional ) , but its performance is very poor with multimodal models like m15 ( watson ) . in its particular comparison with @xmath83 , it can be observed that both selectors share the same order of error , but being @xmath85 better in all the situations except for one : the uniform model ( m1 ) .
this is due to the `` extra term '' commented in section [ kdebwd : sec : ruleofthumb ] : its absence in the denominator makes that @xmath216 faster than @xmath85 when the concentration parameter @xmath217 and , what is a disadvantage for @xmath218 , turns out in an advantage for the uniform case . with respect to @xmath100 and @xmath101 , although their performance becomes more similar when the sample size increases , something expected , @xmath101 seems to be on average a step ahead from @xmath100 , specially for low sample sizes . among the cross - validated selectors , @xmath219 performs better than @xmath73 , a fact that was previously noted by simulation studies carried out by @xcite and @xcite .
finally , @xmath220 presents the most competitive behaviour among the previous proposals in the literature when the sample size is reasonably large ( see table [ kdebwd : tab : rankcirsph ] ) .
+ the comparison between the circular selectors is summarized in the scores of table [ kdebwd : tab : rankcirsph ] .
for all the sample sizes considered , @xmath101 is the most competitive selector , followed by @xmath100 for all the sample sizes except @xmath221 , where @xmath219 is the second .
the effect of the sample effect is also interesting to comment .
for @xmath221 , @xmath219 and @xmath85 perform surprisingly well , in contrast with @xmath220 , which is the second worst selector for this case .
when the sample size increases , @xmath85 and @xmath83 have a decreasing performance and @xmath220 stretches differences with @xmath100 , showing a similar behaviour .
this was something expected as both selectors are based on error criteria that are asymptotically equivalent .
the cross - validated selectors show a stable performance for sample sizes larger than@xmath221 .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] + is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] + is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] + is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] + is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] is drawn .
[ kdebwd : fig : circ],title="fig:",scaledwidth=22.5% ] .comparative study for the circular case , with sample size @xmath222 .
columns of the selector @xmath223 represent the @xmath224 , with bold type for the minimum of the errors .
the standard deviation of the @xmath225 is given between parentheses.[kdebwd : tab : cir ] [ cols=">,>,>,>,>,>,>,>,>,>",options="header " , ] | new bandwidth selectors for kernel density estimation with directional data are presented in this work .
these selectors are based on asymptotic and exact error expressions for the kernel density estimator combined with mixtures of von mises distributions .
the performance of the proposed selectors is investigated in a simulation study and compared with other existing rules for a large variety of directional scenarios , sample sizes and dimensions .
the selector based on the exact error expression turns out to have the best behaviour of the studied selectors for almost all the situations .
this selector is illustrated with real data for the circular and spherical cases .
* keywords : * bandwidth selection ; directional data ; mixtures ; kernel density estimator ; von mises . |
since its discovery in 2004 @xcite , graphene continues to be the subject of intense interest regarding its exotic properties @xcite .
these intriguing properties , such as the anomalous quantum hall effect , are ascribed to dirac type electrons described by the weyl s equation for massless particles @xcite .
the electronic properties in graphene are significantly affected by applying a strain @xcite .
the latter can also , accidentally , occur during the fabrication process as in exfoliation or chemical vapor deposition of graphene samples @xcite .
theoretical and first principle calculations revealed the substantial effect of the strain on the electronic and lattice spectra of graphene @xcite . to bring out the signature of strain induced modified electronic and vibrational properties , raman spectroscopy has emerged as a powerfull probe .
this technique , which is simple to use in graphene , is found to be a successful tool to identify the number of layers in multilayer graphene , to probe the nature of disorder and the doping amount @xcite .
several experimental studies have been carried out on raman spectra of graphene under uniaxial strain @xcite .
the results revealed that , due to the strain , the raman g band is redshifted and splitted into two peaks denoted g@xmath1 and g@xmath2 .
g@xmath1 ( g@xmath2 ) is the mode polarized perpendicular ( along ) the strain direction .
the g peak appearing in unstrained graphene at 1580 @xmath3 corresponds to a doubly degenerate optical mode at the @xmath0 point of the brillouin zone ( bz ) .
the splitting of the g peak results from the strain induced lattice symmetry lowering .
experimental results showed that the frequency shift rates of the g@xmath1 and g@xmath2 as a function of the strain strength @xmath4 is of @xmath5 -13 @xmath6 and @xmath7 -6 @xmath6 @xcite .
recent measurements @xcite reported that the rate shifts of g@xmath2 and @xmath8 are respectively of -33 @xmath6 and -14 @xmath6 in agreement with first principle calculations @xcite .
the difference in the shift rates was attributed to strain calibration @xcite .
the g band splitting could be understood within a phenomenological model based on a semiclassical approach @xcite . within this model , the shear component of the strain
is found to be responsable of the splitting .
raman spectroscopy of strained graphene has also revealed that the 2d band , originating from a resonant scattering process involving two optical phonons at the bz edges , splits into two peaks under uniaxial strain @xcite .
this splitting was ascribed to strain induced changes in the resonant conditions resulting from both modified electronic band structure and phonon dispersion @xcite .
several studies reported that the electron - phonon coupling plays a key role in raman spectroscopy in graphene @xcite .
ando @xcite showed that , in undeformed graphene , the frequency of the center zone optical phonon mode is shifted due to electron - phonon interaction . the frequency behavior is found to depend on the value of the fermi energy @xmath9 compared to the phonon frequency @xmath10 at the @xmath0 point : for @xmath11 ( @xmath12 ) , the phonon frequency is redshifted ( blueshifted ) leading to a lattice softening ( hardening ) . in the clean limit , a logarithmic singularity takes place at @xmath13 which is found to be smeared out in the dirty limit and at finite temperature @xcite .
moreover , ando@xcite reported an anomalous behavior of the optical phonon damping induced by the electron - phonon interaction : for @xmath14 , the phonons are damped due to the formation of electron - hole pairs leading to phonon softening @xcite . however , for @xmath15 , the phonon is no more damped since the electron - hole pair production is forbidden by pauli principle @xcite .
this damping behavior predicted by ando @xcite was observed in raman spectroscopy @xcite . the natural question , which arises at this point , is how the frequency shifts and damping of optical phonon are modified in uniaxial strained graphene where electron band structure is deeply changed .
theoretical studies @xcite showed that the perfect honeycomb lattice of graphene undergoes a quinoid - type deformation by applying a uniaxial strain .
the dirac cones are no longer at the corners of the bz and are tilted .
the corresponding low energy electronic properties could be described by the generalized tow dimensional ( 2d ) weyl s hamiltonian @xcite .
it is worth to note that the tilted dirac cones are also expected in the organic conductor @xmath16-(bedt)@xmath17i@xmath18 where bedt stands for bis(ethylenedithio)-tetrathiafulvalene @xcite.based on the generalized weyl s hamiltonian , several intriguing properties of this compound have been unveiled @xcite . in this paper , we focus on the effect of the electron - phonon interaction on the @xmath0 point optical phonon modes in graphene under uniaxial strain described by a quinoid - type lattice . we show that the frequency shift and the broadening of the longitudinal optical ( lo ) and the transverse optical ( to ) phonon modes are substantially dependent on the characteristic parameters of the weyl hamiltonian which are the tilt and the anisotropy of the electronic dispersion relation .
we bring out original points which , to the best of our knowledge , have not been addressed so far : ( i ) the electron - phonon interaction in strained graphene induces a lifting of the degeneracy of the lo and to modes which contributes to the splitting of the g band .
this effect is found to originate from the anisotropy of the electronic spectrum and not from the tilt of dirac cones .
the latter may only give rise to a global shift of the g band compared to the undeformed case .
the splitting is found to be strongly dependent on the electron density and disorder amount .
( ii ) the anomalous behavior of the phonon damping reported in refs.@xcite in undeformed graphene is found to be a robust feature which is kept under uniaxial strain .
the damping of lo and to modes strongly depends on the strain amplitude and the phonon angle .
we found that , in the particular case , where one of the mode is along the strain direction , the corresponding phonons are strongly damped for a compressive deformation .
however the phonon mode perpendicular to the strain direction is less damped and its lifetime increases as the strain amplitude increases . for tensile deformation
the mode behaviors are exchanged .
( iii ) a crossing of to and lo frequencies can take place at a particular doping values as found in carbon nanotubes @xcite .
( iv ) we found that the electron - phonon interaction contributes to the polarization dependence of the g peak in uniaxial strained graphene as concluded by mohiuddin _
et al._@xcite .
the paper is organized as follows : in sec .
ii we give the outlines of the formulation to derive the optical phonon self - energy .
we start with the generalized weyl s hamiltonian obtained within the effective mass approach .
then , we derive the electron - phonon interaction hamiltonian and the phonon self - energy . the results are discussed in sec .
iii in relation with experiments .
iv is devoted to the concluding remarks .
we consider the optical phonon modes of the center bz responsable of the g peak in graphene .
we focus on the lo and inplane to modes .
we first derive the electronic hamiltonian , within the effective mass theory @xcite , taking into account the first and second neighbor hopping parameters in strained graphene . by applying a uniaxial strain along , for example , the @xmath19 direction
the honeycomb lattice turns to a quinoid type lattice @xcite .
it is worth to note that one should consider an arbitrary strain direction as done for example in refs.@xcite .
however , several experimental and numerical studies @xcite have shown that the g band behavior is independent of the strain direction . considering a generic strain direction will give rise to the same form of the electronic hamiltonian but with renormalized parameters .
we , then , consider for simplicity a strain along the @xmath19 direction as in ref.@xcite .
in such case , the hopping parameter to the first neighboring atoms are no more equal as in undeformed graphene . the distance between neighboring atoms along the @xmath19 direction changes from @xmath20
to @xmath21 the vectors , @xmath22 ( @xmath23 ) , connecting the sites of the a sublattice with first neighbors sites on the b sublattice are given by ( fig.[figlattice ] ) : @xmath24 where @xmath20 is the distance between first neighbor atoms in undeformed graphene , @xmath25 is the lattice deformation which measures the strain amplitude .
@xmath4 is negative ( positive ) for compressive ( tensile ) deformation .
+ the second neighbors sites are connected by vectors @xmath26 given by : @xmath27 where ( @xmath28 ) is the lattice basis . the hopping integral along @xmath29
is affected by the strain and is different from those along @xmath30 and @xmath31 which are equal .
moreover , the hopping parameters to the second neighboring atoms along @xmath32 and @xmath33 are modified by the strain compared to that along @xmath34 .
it is worth to stress that by applying a strain along the @xmath19 direction one should expect a strain component along the @xmath35 axis @xmath36 where @xmath37 is the poisson ratio of graphene .
the off diagonal terms of the strain tensor , which depend on the strain direction and the poisson ratio@xcite , generate different bond lengths .
however , for a strain axis parallel to the principal symmetry direction @xmath35 or @xmath19 , these terms vanish leading to equal bond lengths as assumed in our model .
the contribution of poisson ration could , then , be neglected compared to the main contribution resulting from the strain component along the stress axis . axis .
( @xmath28 ) is the lattice basis .
the hopping parameters to the first ( second ) neighbors @xmath38 and @xmath39 ( @xmath40 and @xmath41 ) are different due the deformation . vectors connecting first ( second ) neighboring atoms are denoted @xmath22 ( @xmath26 ) . ]
we denote by @xmath42 ( @xmath43 ) the hopping integral to the first ( second ) neighboring atoms along @xmath44 ( @xmath26 ) vectors .
we set @xmath45.under strain @xmath46 changes from @xmath38 to @xmath39 given by @xcite @xmath47 @xmath43 along @xmath32 and @xmath33 changes from the value of undeformed graphene , denoted @xmath40 , to @xmath48 written as : @xmath49 the momentum vectors of dirac points @xmath50 and @xmath51 are given respectively by @xcite @xmath52 where @xmath53 is the valley index .
we denote hereafter @xmath54 in undeformed graphene , the dirac points @xmath50 and @xmath51 are at the corners of the bz @xmath55 and @xmath56 . under the strain , @xmath50 and @xmath51
move away from @xmath55 and @xmath56 points @xcite .
+ the electronic wave function can be written as @xcite : @xmath57 where @xmath58 and @xmath59 are atomic orbitals centred on atoms a and b respectively .
+ in the @xmath60 approach @xcite , the coefficients @xmath61 and @xmath62 are given by : @xmath63 where @xmath64 and @xmath65 are slowly varying envelope functions . + considering second neighbor hopping integrals , the electronic energy obeys to : @xmath66 where @xmath67 , @xmath68 and @xmath69 .
+ within the @xmath60 method , eq.[eigen ] becomes : @xmath70 where @xmath71 is the wave vector and @xmath72 details of the calculations are given in appendix a. + from eq.[eigen ] we recover the so - called minimal form of the generalized weyl hamiltonian @xcite : @xmath73 where @xmath74 , @xmath75 , @xmath76 and @xmath77 are the 2x2 pauli matrices .
the corresponding dispersion relation is of the form : @xmath78 @xmath79 is responsable of the tilt of dirac cones away from the @xmath80 axis .
this term obeys to the condition @xcite @xmath81 which insures the presence of two energy bands : a positive energy for @xmath82 and a negative energy band for @xmath83 @xcite . in deformed graphene and for @xmath84 , @xmath85 @xcite .
the eigenfunctions of the hamiltonian given by eq.[helec ] are of the form : @xmath86 where @xmath87 is the chirality index , @xmath88 is the lattice surface under strain and @xmath89 . in this section ,
we derive the effective hamiltonian describing the effect of the lattice vibrations on the electronic hamiltonian .
such effect arises from the change of the hopping integrals due to the lattice distortion .
this hamiltonian was obtained by ando @xcite in the case of undeformed graphene .
we shall determine the electron - phonon interaction hamiltonian in quinoid - type deformed graphene .
+ the phonon hamiltonian can be written as @xcite @xmath90 where @xmath91 ( @xmath92 ) is the creation ( annihilation ) operator of phonon with wave vector @xmath93 and mode @xmath94 lo , to .
@xmath95 is the @xmath96 mode phonon frequency at the @xmath0 point . +
the relative displacement of the two sublattices a and b in the continuum limit is @xmath97 which can be written for optical phonon at @xmath0 point as @xcite : @xmath98 where @xmath99 is the mass of the carbon atom , @xmath100 is the number of unit cells and @xmath101 is given by : @xmath102 with @xmath103 .
+ to derive the electron - phonon effective hamiltonian , we shall determine the effect of the lattice displacement on the hopping integrals .
+ the hopping parameter between first neighboring atoms located at @xmath104 and @xmath105 is changed from @xmath42 to @xcite : @xmath106 \label{tnn}\end{aligned}\ ] ] with @xmath107 , @xmath108 and @xmath109 .
the hopping integral between second neighboring atoms changes from @xmath43 to @xmath110 \label{tnnn}\end{aligned}\ ] ] however , the correction to @xmath43 terms vanishes for @xmath0 point optical phonon modes ( @xmath111 ) .
+ since the amplitude of the lattice displacement is small compared to the lattice parameter , eq.[tnn ] becomes : @xmath112 \label{tnncorrect}\end{aligned}\ ] ] in the continuum limit , @xmath113 .
the correction to the hopping integrals due to lattice distortion , given by eq.[tnncorrect ] , leads to an extra term @xmath114 in the electronic hamiltonian which is written near the d point as ( for details , see appendix b ) : @xmath115 where @xmath116 and @xmath117 are the component of the relative displacement @xmath118 .
@xmath119 is of the form : @xmath120\sim w_y-2\epsilon t^{\prime } a ( 1+\epsilon)\end{aligned}\ ] ] and @xmath121 obeys to eq.[theta ] .
+ given the expression of @xmath122 and since @xmath123 @xcite , we have @xmath124 and @xmath125 with @xmath126 . + the electron - phonon hamiltonian can , then , be written as @xcite : @xmath127 where @xmath128 , @xmath129 is the frequency of he optical phonon at @xmath0 point in the deformed graphene for the mode @xmath96 in the absence of electron - phonon interaction . in undeformed graphene @xmath130
this degeneracy is expected to be lifted in the strained graphene due to the symmetry breaking .
according to a phenomenological model @xcite the strain tension @xmath131 in graphene reduces to @xmath132 where @xmath19 is the direction of the applied strain , and @xmath133 along the direction transverse to the strain and @xmath134 is the poisson ratio .
the g band splits into two bands @xmath135 with frequencies @xmath136 shifted from the unstrained band frequency @xmath137 as @xmath138 where @xmath139 and @xmath140 are respectively the grneisen parameter and the shear deformation potential .
the shear component of the strain , @xmath141 , is then responsible of the g band splitting .
the question arising at this point concerns the contribution of the electron - phonon interaction to the splitting of the g band . to highlight this contribution
, we did not consider the effect of the shear component which turns out to disregard the effect of the strain on the phonon dispersion .
we then assume that , in the absence of electron - phonon interaction , the center zone optical phonon modes lo and to have the same frequencies @xmath142 . by switching on the interaction , this degeneracy may be lifted giving rise to two bands corresponding to the lo and to modes which results in the g band splitting .
+ the matrices @xmath143 are given , near d point , by : @xmath144 where @xmath145 .
@xmath143 near @xmath51 point satisfies @xmath146 @xcite.contrary to acoustic phonons , there is no scalar deformation potential in the interaction hamiltonian @xcite regarding the expression of the relative displacement of the long wavelength optical phonons ( eq.[displa ] ) .
the retarded phonon green function can be written as @xcite @xmath147 @xmath148 is the self - energy and @xmath149 , @xmath150 being the scattering time .
+ the shift @xmath151 of the phonon frequency is given by the real part of the green function s pole .
for small correction to @xmath10 , @xmath152 is given by : @xmath153 the imaginary part of the green function s pole gives the broadening @xmath154 of the phonon mode .
@xmath155 being the phonon lifetime : @xmath156 the self - energy of @xmath0 point optical phonon can be written as @xcite @xmath157 where @xmath158 and @xmath159 are the valley and spin degeneracy , @xmath160 is the fermi distribution function @xmath161 and @xmath162 is the chemical potential at temperature @xmath163 .
@xmath88 is the graphene surface under uniaxial strain @xmath164 where @xmath165 is the undeformed graphene surface . for long wavelength phonon modes near @xmath50 point
, the matrix elements can be written as : @xmath166\nonumber\\ % \frac{w_xw_y^{\prime}}2 \left[\alpha^{\prime2}\sin^2\varphi(\vec{q})\left(1+\cos2\phi_{\vec{k}}\right)+ \frac{\cos^2 \varphi(\vec{q})}{\alpha^{\prime 2}}\left(1-\cos2\phi_{\vec{k}}\right ) -\sin2\varphi(\vec{q})\cos2\phi_{\vec{k}}\right]\nonumber\\\end{aligned}\ ] ] according to eq.[self ] only interband processes ( @xmath167 ) contribute the self - energy of @xmath111 phonon modes . regarding the electronic dispersion relation ( eq.[dispers ] ) , the term @xmath168 , in eq.[self ] , becomes @xmath169 setting @xmath170 and @xmath171 , the integration over @xmath172 in eq.[self ] vanishes and the expression of the self - energy can be reduced to an integration over the energy : @xmath173\left[\frac 1 { \hbar \omega+2\varepsilon+i\eta}- \frac 1{\hbar \omega-2\varepsilon+i\eta}\right ] \label{self3}\end{aligned}\ ] ] where we used the density of state in quinoid lattice @xmath174 @xcite .
@xmath175 is a renormalized fermi velocity given by @xcite @xmath176 @xmath177 in eq.[self3 ] is a cutoff energy corresponding to the limit of validity of the linear electronic dispersion given by eq.[dispers ] and the coefficient @xmath178 is given by : @xmath179\nonumber\\ c_t&=&a\left[\alpha^{\prime 2}\sin^2\varphi(\vec{q})+\frac{\cos^2\varphi(\vec{q})}{\alpha^{\prime 2}}\right]\end{aligned}\ ] ] and @xmath180 is a constant written as : @xmath181 as mentioned in ref.@xcite , one should substract the contribution of @xmath182 modes to avoid double counting of electron contribution .
the self - energy at zero temperature takes , then , the form : @xmath183 \left[ae_f^{\ast}-\frac 14 a(\hbar \omega + i\eta)\ln\left(\frac{\hbar\omega+2e_f^{\ast}+i\eta } { \hbar\omega-2e_f^{\ast}+i\eta}\right)+i\pi\right]\nonumber\\ \pi_{t}(\vec{q}\rightarrow \vec{0},\omega)=\frac 1{\left(1-\tilde{w}_0 ^ 2\right)^{\frac 32}}\left[\alpha^{\prime 2}\sin^2\phi + \frac{\cos^2\phi}{\alpha^{\prime 2}}\right ] \left[ae_f^{\ast}-\frac 14 a(\hbar \omega + i\eta)\ln\left(\frac{\hbar\omega+2e_f^{\ast}+i\eta } { \hbar\omega-2e_f^{\ast}+i\eta}\right)+i\pi\right]\nonumber\\ \label{self5}\end{aligned}\ ] ] where we set @xmath184 and @xmath185 ( see appendix a ) , with @xmath186 being the fermi energy in undeformed graphene . eq.[self5 ] reduces to that obtained by ando @xcite in undeformed graphene for @xmath187 and @xmath188 .
figures [ shift02 ] shows the dependence of the frequency shifts and broadening of the lo and the to modes as a function of the fermi energy @xmath9 in the dirty limit for a compressive strain strength @xmath189 .
the shifts are normalized to @xmath190 where @xmath180 is given by eq.[a].in undoped system , the effect of electron - phonon interaction on the frequency shifts is not relevant .
this effect is enhanced by introducing impurities in the system or by increasing the strain amplitude as we will show in the next section . + in the dirty limit @xmath191 for a compressive strain @xmath192 .
the lo mode is along the strain axis .
the solid line is the result for the undeformed case.,title="fig : " ] + ( a ) in the dirty limit @xmath191 for a compressive strain @xmath192 .
the lo mode is along the strain axis .
the solid line is the result for the undeformed case.,title="fig : " ] + ( b ) for clarity reasons , we will consider in the following strain strength @xmath193 .
it should be noted that the critical strain for graphene is of @xmath194.figure [ shift0 ] shows the dependence of the frequency shifts on the fermi energy @xmath9 in the clean limit ( @xmath195 ) for a compressive strain strength @xmath196 .
+ due to the deformation , the degeneracy of lo and to modes , obtained in the undeformed graphene ( solid line in fig.[shift0 ] ) , is lifted.the logarithmic singularity at @xmath197 reported in the undeformed case is a robust feature which persists under strain but takes place at @xmath198 which corresponds to @xmath199 in eq.[self5 ] . in the clean limit ( @xmath195 ) and for a compressive strain @xmath200 .
the lo mode is along the strain axis .
the solid line is the result for the undeformed case .
the inset shows the frequency shifts for @xmath201 . ]
according to fig.[shift0 ] , both to and lo modes are redshifted leading to a lattice softening for @xmath202 .
however , the phonon frequencies increase with @xmath9 and the lattice hardens for @xmath203 . moreover , the frequency of the lo mode , which is along the strain axis , is more shifted compared the the to mode .
the lo mode is , then , more affected by the electron - phonon interaction as shown by the broadening behavior depicted in figure [ broad0 ] .
the damping of the lo mode is more pronounced than that of the to mode which is found to be more long lived than the modes of undeformed graphene . in the clean limit ( @xmath195 ) and for a compressive strain @xmath200 .
the lo mode is along the strain axis .
the solid line is the result for the undeformed case . ]
this behavior can be understood from the structure of the electronic dispersion . along the strain direction ,
the electron velocity is enhanced for a compressive deformation ( @xmath204 ) as @xmath205 , while that in the perpendicular direction is reduced as @xmath206 .
the fermi level changes as @xmath207 which increases for a compressive strain ( fig.[pauli ] ) . as a consequence ,
the production of electron - hole pairs is furthered along the strain direction , as shown in figure [ pauli ] , since there are more states which are not blocked by pauli principle for a given phonon frequency .
however , in the direction perpendicular to the strain , electron - hole processes , allowed in the undeformed case , become forbidden by the pauli exclusion principle .
this explains the long lived to phonon mode compared to the modes of undeformed graphene .
+ ) increases under compressive deformation and the fermi velocity @xmath208 ( @xmath209 ) along ( perpendicular ) to the strain direction @xmath210 ) is enhanced ( reduced ) compared to isotropic case .
this leads to more ( less ) electron - hole pairs contributing to phonon softening .
the dashed and solid arrows ( crossed solid and dashed arrows ) denote the electron - hole process leading to phonon softening ( hardening ) for undeformed and compressed case respectively.,title="fig : " ] ) increases under compressive deformation and the fermi velocity @xmath208 ( @xmath209 ) along ( perpendicular ) to the strain direction @xmath210 ) is enhanced ( reduced ) compared to isotropic case .
this leads to more ( less ) electron - hole pairs contributing to phonon softening .
the dashed and solid arrows ( crossed solid and dashed arrows ) denote the electron - hole process leading to phonon softening ( hardening ) for undeformed and compressed case respectively.,title="fig : " ] the behavior of lo and to modes are exchanged for @xmath211 where the to mode becomes along the strain direction .
moreover , the behavior are also exchanged for tensile deformation ( @xmath212 ) . +
this feature can be understood from eq.[self5 ] showing that the leading term for the frequency shifts is @xmath213 in compressive strain and @xmath214 for tensile deformation .
+ figure [ shiftphi ] shows the frequency shifts and the broadening of the phonon modes at @xmath215 .
the difference in damping of to and lo modes , obtained for @xmath216 and @xmath211 , is clearly reduced since both modes have a component along the strain direction . in the clean limit ( @xmath195 ) and for a compressive deformation @xmath200 .
the phonon angle is @xmath215 .
the solid line is the result for the undeformed case.,title="fig : " ] + ( a ) in the clean limit ( @xmath195 ) and for a compressive deformation @xmath200 .
the phonon angle is @xmath215 .
the solid line is the result for the undeformed case.,title="fig : " ] + ( b ) the logarithmic singularity obtained in the clean limit at @xmath199 ( fig.[shift0 ] ) is smeared out in the dirty limit as shown in fig.[loga ] for @xmath216 in the case of tensile and compressive deformation . according to fig.[loga ] ,
the frequency shifts of lo and to modes depend on the fermi level and the amount of disorder . away from @xmath217 ,
all modes show a blueshift contrary to the clean limit where lo and to modes undergo a redshift ( blueshift ) for compressive ( tensile ) strain at @xmath218 .
the frequency blueshift is reminiscent of that found by ando@xcite in undeformed graphene in the dirty limit .
+ the dependence of the frequency shifts on the doping level and the amount of disorder may explain the discrepancy in the experimental values of the shift rates of g@xmath1 and g@xmath2 bands as function of the strain @xcite and which was ascribed to a difference in the strain calibration .
we suggest that , this discrepancy may be due to the doping and the disorder amount in the sample .
+ in ref.@xcite , the authors studied the behavior of the g band in deformed graphene using polarized light .
they reported that the g peak can be regarded as mixture of three peaks corresponding to undeformed case ( g@xmath219 ) , compressive ( g@xmath2 ) and tensile ( g@xmath1 ) deformation .
the authors attributed the presence of both blue and red shifted frequencies ( g@xmath1 and g@xmath2 bands ) to the anisotropy of the applied deformation .
according to figs.[shift0 ] and [ loga ] , for @xmath220 and @xmath216 , the lo mode ( to mode ) is blueshifted ( redshifted ) compared to the undeformed mode ( solid line in the figures ) for compressive strain .
the experimental results of ref.@xcite could then be the signature of the electron - phonon interaction .
the shifted g@xmath1 and g@xmath2 modes could be assigned to the lo and to modes for a given uniaxial strain at a doping level @xmath220 . + in the dirty limit ( @xmath221 ) and for compressive strains of @xmath222 and @xmath223 .
the lo phonon mode is along the strain direction .
the solid line is the result for the undeformed case . ] in figure [ broad ] , we plot the broadening of phonon modes as a function of the fermi energy for @xmath216 in the dirty limit .
the figure shows that the damping of the mode along the strain direction is enhanced as the amplitude of the deformation increases .
this reflects the increasing number of the electron - hole pairs leading to decaying phonons ( fig.[pauli ] ) . in the dirty limit ( @xmath221 ) and for compressive strains of @xmath222 and @xmath223 .
the lo phonon mode is along the strain direction .
the solid line is the result for the undeformed case . ]
the strain dependence of the frequency shifts is depicted in fig.[shiftstrain ] where we considered the case of undoped graphene in the dirty limit ( @xmath224 , @xmath225 ) and the doped graphene ( @xmath226 ) in the clean limit since the shifts in the clean undoped case are small .
the shift behaviors could be understood from the processes depicted in fig.[pauli ] . ) , ( b ) in doped case ( @xmath226 ) and in the clean limit ( @xmath195 ) and ( c ) in doped case ( @xmath227 ) and in the dirty limit ( @xmath221 . the lo phonon mode is along the strain direction.,title="fig : " ] + ( a ) ) , ( b ) in doped case ( @xmath226 ) and in the clean limit ( @xmath195 ) and ( c ) in doped case ( @xmath227 ) and in the dirty limit ( @xmath221 . the lo phonon mode is along the strain direction.,title="fig : " ] + ( b ) ) , ( b ) in doped case ( @xmath226 ) and in the clean limit ( @xmath195 ) and ( c ) in doped case ( @xmath227 ) and in the dirty limit ( @xmath221 . the lo phonon mode is along the strain direction.,title="fig : " ] + ( c ) fig.[shiftstrain ] shows a linear behavior of the frequency shift as a function of the strain strength for small strain .
this is reminiscent of the experimental results reported in refs .
@xcite . the strain rates and slopes of the frequency shifts are dependent on the doping level and the disorder amount . according fig.[shiftstrain ] ,
the linearity is lost by increasing the strain .
it is worth to note that a departure from a linear behavior was also reported in ref.@xcite for the strain dependence of the frequency shift of the 2d raman band .
such behavior could also be observed in raman spectra of @xmath16(bedt)@xmath17i@xmath17 salt showing a strong anisotropic electronic dirac spectrum .
+ in the limit of strong strain , we expect a decoupling of electron - hole pairs from the phonon mode along ( perpendicular ) to the strain axis for tensile ( compressive ) deformation as shown in fig.[shift4 ] .
such effect could not be observed in graphene where the critical strain is of 25@xmath228 but may be bring out in @xmath16(bedt)i@xmath17 @xcite .
+ in the dirty limit ( @xmath221 ) and for a strong tensile deformation .
the lo phonon mode , which is along the strain direction , decouples from the electron - hole pairs.,title="fig : " ] + ( a ) in the dirty limit ( @xmath221 ) and for a strong tensile deformation .
the lo phonon mode , which is along the strain direction , decouples from the electron - hole pairs.,title="fig : " ] + ( b ) a hallmark feature of the doping dependence of the frequency shifts is the presence of crossings of lo and to modes ( figs.[shift0 ] , [ loga ] ) . at the corresponding fermi energy , no g band splitting
is expected due to electron - phonon interaction .
experimentally , the g@xmath1 and the g@xmath2 bands should then merge in uniaxial strained graphene by doping the sample at the critical value corresponding to the crossing of lo and to modes .
this feature could only be observed in the absence of the shear strain which induces a splitting of the g band . a possible
crossing of lo and to modes was also reported in carbon nanotubes @xcite .
+ in figure [ phi ] , we plot the dependence of the phonon frequency shifts on the phonon angle @xmath229 with respect to the axis perpendicular to the strain direction .
the shifts of the lo and to modes display a periodic modulation with a relative shift of 90@xmath230 . according to eq.[self5 ] , this dependence is due to the anisotropy of the electronic dispersion relation . considering the isotropic case ( @xmath231=1 ) , the shifts become independent on @xmath229 as in isotropic honeycomb lattice @xcite . ) , in the dirty limit ( @xmath221 ) and for a compressive deformation .
@xmath229 ( in degree ) is the angle of optical phonon with respect to @xmath35 axis perpendicular to the strain direction .
the solid line is the result for the undeformed case ] our results are in agreement with the experimental data @xcite and numerical calculations @xcite showing a periodic modulation of the intensity of g@xmath1 and the g@xmath2 peaks as a function of the angle between the incident light polarization and the strain axis .
the relative shifts of the two bands is also found to be of 90@xmath230 .
our results support the idea presented in the experimental study of mohiuddin _ _ et al.__@xcite suggesting that the polarization dependence of the g peaks is due to the anisotropy of the electronic spectrum and such dependence is the signature of the electron - phonon intercation .
it is worth to stress that sasaki _
et al._@xcite proposed that the nature of the graphene edges contributes also to the polarization dependence of raman bands in strained graphene .
we have derived the frequency shifts and the broadenings of the longitudinal ( lo ) and transverse ( to ) optical phonon modes at @xmath0 point in graphene under uniaxial strain disregarding the contribution of shear strain component .
we show that the raman g band , corresponding to a double degenerate mode in undeformed graphene , may split into two peaks due to electron - phonon interaction .
these peaks are assigned to the lo and the to modes which are found to be strongly dependent on the fermi level and the amount of disorder .
this dependence may explain the difference in the experimental results giving the strain rates of the frequency shifts of the g@xmath1 and g@xmath2 modes .
moreover , we found that the splitting of the g band is due to the anisotropy of the electronic spectrum .
the tilt of dirac cones , arising also from the strain , is found to be irrelevant for the relative frequency shift of the lo and to modes since it leads to a global shift of the g peak .
we also show that the electron - phonon intercation contributes to the raman polarization dependence of the g peaks in strained graphene .
this contribution reflects the anisotropy of the electronic spectrum .
the optical phonon mode along the strain is found to be damped ( long lived ) for compressive ( tensile ) strain .
the frequency shifts and the lifetime of the optical phonons are substantially dependent on the strain strength and the phonon angle . at relatively strong strain , it is possible to induce a decoupling of the phonon mode perpendicular to the compressive strain axis from electron - hole pair production process .
the signature of the strain induced anisotropic electronic dispersion could also be brought out in the @xmath0 point magnetophonon resonance at high magnetic field @xcite .
we warmly thank y .- w .
son , k. sasaki for helpful and stimulating discussions .
we thank j .- n .
fuchs for a critical reading of the manuscript .
this work was partially supported by the national research foundation of korea ( nrf ) grant funded by the korea government ( mest ) ( no .
2011 - 0030902 ) .
s. h acknowledges the kind hospitality of w. kang and the members of cerc ( seoul , korea ) .
the final form of the manuscript was prepared in the ictp ( trieste , italy ) .
s. h. was supported by simons - ictp associate fellowship .
the @xmath60 method was used by ando @xcite to derive the dirac hamiltonian in undeformed graphene taking only into account the hopping integral to the first neighboring carbon atoms .
following ref.@xcite , we derive the weyl hamiltonian for uniaxial strained graphene considering first and second neighboring hopping integrals .
we start with the eigenproblem given by eq.[eigen ] where the functions @xmath61 and @xmath232 are written as : @xmath233 here the vectors @xmath234 , @xmath235 , @xmath236 and @xmath237 are given by : @xmath238 the l.h.s of eq.[eigen ] can be written , at @xmath104 as : @xmath239 where @xmath240 is a smoothing function satisfying : @xmath241 @xmath242 is an envelope function@xcite .
these properties yield to @xmath243 which is reminiscent of the @xmath244 function @xcite .
eq.[eigen ] can then be written , around the a site , as : @xmath245 the l.h.s of eq.[eigen2 ] reduces to @xmath246 and , in the r.h.s , we set : @xmath247 we then obtain @xmath248 applying this term to @xmath249 in eq.[eigen2 ] and summing over @xmath250 gives rise to a diagonal term of the form @xmath251 which leads to a shift of the total energy .
the term @xmath252 in eq.[dl ] , summed over @xmath250 and applied to @xmath249 , gives : @xmath253 in eq.[eigen2 ] , the contribution of the first neighbor hopping integrals gives then rise to the following eigenproblem near @xmath50 point : @xmath254 where @xmath255 , @xmath256 , @xmath257 and @xmath258 .
+ for the second neighbor hopping integrals , one have : @xmath259 and @xmath260 where @xmath261 .
+ the electronic hamiltonian , near @xmath50 and @xmath51 points , takes the form : @xmath262 with @xmath263 ( - ) at @xmath50 ( @xmath51 ) point . @xmath264 and
@xmath122 can be expressed as a function of the strain strength as @xmath265 in graphene , @xmath266 @xcite . in the present case
, we have @xmath84 . +
regarding the effect of the lattice distortion on the hopping integral ( eq.[tnn ] ) an extra term appears in the electronic hamiltonian ( eq.[weyl ] ) .
this term arises from the contribution of the hopping term correction @xmath267 in eq.[eigen ] .
this contribution is of the form k. s. novoselov , a. k. geim , s. v. morozov , d .
jiang , y. zhang , s. v. dubonos , i. v. grigorieva , and a. a. firsov , science , * 306 * 666 ( 2004 ) , k. s. novoselov , a. k. geim , s. v. morozov , d. jiang , m. i. katsnelson , i. v. gregorieva , s. v. dubonos , and a. a. firsov , nature * 438 * , 197 ( 2005 ) . t. m. g. mohiuddin , a. lombardo , r. r. nair , a. bonetti , g. savini , r. jalil , n. bonini , d. m. basko , c. galiotis , n. marzari , k. s. novoselov , a. k. geim , and a. c. ferrari , phys .
b * 79 * , 205433 ( 2009 ) .
o. frank , g. tsoukleri , i. riaz , k. papagelis , j. parthenios , a. c. ferrari , a. k. geim , k. s. novoselov and c. galiotis , nature communications * 2 * , 255 ( 2011 ) . o. frank , m. mohr , j. maultzsch , c. thomsen , i. riaz , r. jalil , k. s. novoselov , g. tsoukleri , j. parthenios , k. papagelis , l. kavan , and c. galiotis , acs nano , * 5 * 2231 ( 2011 ) . | we derive the frequency shifts and the broadening of @xmath0 point longitudinal optical ( lo ) and transverse optical ( to ) phonon modes , due to electron - phonon interaction , in graphene under uniaxial strain as a function of the electron density and the disorder amount .
we show that , in the absence of a shear strain component , such interaction gives rise to a lifting of the degeneracy of the lo and to modes which contributes to the splitting of the g raman band .
the anisotropy of the electronic spectrum , induced by the strain , results in a polarization dependence of the lo and to modes .
this dependence is in agreement with the experimental results showing a periodic modulation of the raman intensity of the splitted g peak .
moreover , the anomalous behavior of the frequency shift reported in undeformed graphene is found to be robust under strain . |
the subject of this work is mathematical modeling of state - of - charge in simple battery cells , such as a non - rechargeable 3 volts lithium coin battery .
the goal is to understand the response of the battery , and ultimately to predict battery lifetime , as energy is consumed under a given discharge usage pattern .
the main incentive for our work is the battery usage in wireless sensor networks and similar internet - of - things systems .
these networks consist of inter - connected low - cost nodes , equipped with basal sensors , computer , radio and a battery , expected to run for many years under very low intensity loads and short dutycycles . within the tight cost restrictions typical of such systems , methods or techniques providing `` battery - charge indicators ''
do not seem to be within reach currently . to make progress in this direction
it is essential to address the problem of predicting battery life .
our paper intends to cover some of the required modeling groundwork and discuss analysis of linear , nonlinear and stochastic aspects of modeling battery cells .
mathematical modeling of batteries has developed over several decades along with the growth of new battery technologies and materials . yet
, there has been relatively little in - depth study of widely available , inexpensive coin cell batteries and on special load characteristics including short load periods .
primarily , lithium and lithium - ion battery models have been developed within electrochemical engineering .
two recent survey and review works @xcite represent the state - of - art of modeling based on the fundamental principles of electrochemistry , and emphasize the wide range of scales involved .
the temporal and spatial scales of the physics and chemistry of the battery range from macroscopic level all the way down to the atomistic level . the tutorial review by landstorfer and jacob @xcite provides a framework of non - equilibrium thermodynamics as the foundation for studying the electrode , electrolyte and interface reactions in great detail .
the review work by ramadesigan _
@xcite , summarizes the literature on such models and , in addition , brings a systems engineering approach applied to li - ion batteries .
this type of model can be said to begin with the pseudo - two - dimensional ( p2d ) model of doyle _
@xcite , which leads to a coupled system of non - linear pdes . more generally , coupled systems of equations with complex boundary conditions are derived , which connect charge concentrations with transport and kinetics of reactant species .
several approaches have been proposed to simplify the resulting sets of equations and allow for numerical computations , see e.g. @xcite . in this work
we use the different mathematical approach to battery modeling developed in communication engineering for computer science applications , see e.g. jongerden and haverkort @xcite . where chemical engineering modeling typically begins with a detailed scheme of reactions and mechanisms in the various phases and interfaces of the cell , these models view the battery as a generic device subject to some fundamental principles .
the main focus of the modeling changes and is now rather the response of the battery to external load .
a typical purpose is load scheduling to optimize battery utilization .
important aspects of battery behavior from this point of view are the rate - capacity effect and charge recovery .
quoting @xcite : _ the former refers to the fact that a lower discharge rate is more efficient than a higher ; more charge can be extracted from the battery before reaching a given cut - off value .
the latter refers to the fact that an intermittent discharge is more efficient than a continuous one . because of these effects , different battery loads that use the same total charge do not result in the same device lifetime . _
the most basic of these methods use linear odes @xcite and gradually build complexity by using pdes and other means @xcite .
a further direction is stochastic modeling using markov chain dynamics @xcite .
the aim of our work is to investigate to what extent these simplified battery models are able to capture important aspects of battery behavior , and to get new insights by developing the mathematical models further .
of special interest is the essential response of the battery to deterministic or random on - off discharge patterns typical for batteries in wireless sensor networks , and the ability of a cell to recover charge during operation .
charge recovery is believed to depend on a number of internal mechanisms , such as convection , diffusion , migration , and charge - transfer . in the modeling work we put special emphasis on separating the roles of diffusion , which is the motion of electroactive species in the presence of a concentration gradient , and migration , the motion of charged species in an electric field .
our starting point is the kinetic battery model of manwell and mcgowen @xcite , which describes the joint evolution of available charge and bound charge over time .
charge recovery in this framework consists in the continuous transition of bound charge to available charge .
as observed in @xcite and further investigated in @xcite , more general spatial versions of these models are related to the diffusion model of rakhmatov and vrudhula @xcite , and leads to a class of second order diffusion equations with robin type boundary conditions .
quoting @xcite , the linear dynamics of the kinetic battery model is _ useful in getting an intuitive idea of how and why the recovery occurs but it needs a number of additions to be useful for the types of batteries used in mobile computing_. using experimental data from ni - mh batteries , @xcite proposed a modified , non - linear , factor in the flow charge and discussed related stochastic versions of the model . in an effort to compare more systematically linear and nonlinear dynamics in kinetic and stochastic battery models , we study discrete time markov chains with nonlinear jump probabilities derived from a simplified charge transport scenario . by a scaling approximation
we obtain in the limit a deterministic nonlinear ode , with explicit solutions which , in principle , can be compared to those of the linear approach .
the unifying mathematical aspect in our analysis is the representation of remaining and available capacities as time - autonomous systems . in section 2 ,
following preliminaries on capacities , internal charge recovery and discharge profiles , we consider the kinetic battery model and discuss a variation .
then we set up an extended version of the spatial kinetic battery model with a finite number of serial compartments , derive the spatially continuous limiting pde , and give a probabilistic representation of the solution of the pde in terms of brownian motion with drift reflected at the boundaries on both sides of a finite interval .
the solution represents the capacity storage of a battery and these tools allow us to study the balance of available and remaining stored capacity .
section 3 studies a nonlinear , stochastic markov chain model and its deterministic ode approximation . with proper choice of nonlinear dynamics for charge recovery due to transfer , diffusion and migration effects
, we then propose a somewhat wider class of nonlinear ode of potential use for battery modeling . as a consequence ,
it is possible to study performance measures such as battery life , delivered capacity and gain , and to compare and optimize the performance of batteries .
we consider a primary ( non - rechargeable ) battery cell consisting of two electrodes , anode and cathode , linked by an electrolyte .
the cell contains a certain amount of chemically reactive material which is converted into electrical energy by an oxidation reaction at the anode .
primary lithium batteries have a lithium anode and may have soluble or solid electrolytes and cathodes .
the mass of material involved in the battery reaction yields a higher concentration of electrons at the anode , and hence by faraday s first law the transfer of a proportional quantity of electrical charge .
this determines a terminal voltage between the pair of electrodes . by closing
a wired circuit between the terminals a current of electrons will start moving through the wire from anode to cathode where they react with a positively charged reactant , manifesting the ability of the battery to drive electric current .
the intensity of the current depends on the total resistance along the wire . inside the battery
the movement of charge - carriers forms a corresponding ionic current , which is controlled by a variety of mechanisms , among them migration of ions , diffusion of reactant species , and charge - transfer reaction .
migration is generated by an electric potential gradient ( electric field ) and convective diffusion by the concentration gradient .
conductivity arises from the combination of migration and diffusion .
charge - transfer reactions take place when migrating ions are transferred from the electrolyte bulk through the anode surface .
it is sometimes helpful to keep track of proper units .
the battery has a given voltage @xmath0 in volts [ v ] and a theoretical capacity @xmath1 in ampere hours [ ah ] , representing the entire storage of chemically reactive material in the cell .
we write @xmath2 [ ah ] for the nominal capacity of a fully loaded cell , where @xmath3 .
this is the amount of electric charge which is delivered if the cell is put under constant , high load and drained until a predefined cut - off voltage is reached .
measuring time @xmath4 in hours [ h ] we write @xmath5 for the consumed capacity [ ah ] and @xmath6 for the remaining capacity [ ah ] , at time @xmath4 . here , @xmath7 is an increasing function , typically continuous with the slope representing intensity of the current .
it is also plausible to let the discharge function have jumps to be interpreted as spikes of charge units being released point - wise to the device driven by the battery . as
an additional level of generality it is straightforward to consider @xmath8 defined on a probability space and subject to a suitable distributional law of a random process .
the instantaneous discharge current [ a ] at @xmath4 is the derivative @xmath9 and the average discharge current [ a ] is the quantity @xmath10 , assuming this limit exists .
we are interested in the behavior of the battery cell when exposed to the accumulated discharge process @xmath8 , in particular regarding @xmath11 at time $ t$ } , \quad u(0)=n\\[1 mm ] \widetilde u(t)&=u(t)/n= \mbox{state of charge at time $ t$},\quad \widetilde u(0)=1\\[1 mm ] e(t)&=\mbox{voltage [ v ] at time $ t$},\quad e(0)=e_0 .
\end{aligned}\ ] ] while there is no obvious method of observing state - of - charge empirically , voltage is accessible to measurements at least in principle . to describe typical voltage
, we imagine that a fully charged battery at time @xmath12 is connected to a closed circuit at constant discharge current @xmath13 [ a ] .
the result is an instant voltage drop from @xmath14 to a new level at approximate voltage @xmath15 , where @xmath16 is an internal resistance [ ohm ] of the cell .
as long as charge is consumed , the state - of - charge will then begin to decline over time accompanied by a subsequent change of voltage .
if after a period of discharge the current is disconnected and the battery temporarily put to rest , then the voltage increases .
first , instantly , by the amount @xmath17 and then over the course of the off - period at some rate due to recovery effects inside the cell
. the resulting voltage versus time curve extended over a longer time span would typically stay nearly constant or exhibit slow decline over most of the active life of the battery followed by a steeper decent until a cut - off level @xmath18 is reached , beyond which the cell is considered to be non - operational . to relate state of charge and voltage we recall that the actual load a time @xmath4 is given by @xmath19 and apply the equilibrium nernst equation , @xcite ch . 1
, to obtain @xmath20 where @xmath21 is the ideal gas constant , @xmath22 is absolute temperature , @xmath23 is the valency of the battery reactant ( @xmath24 for lithium ) , @xmath25 is faraday s constant , and dimensions are such that @xmath26 is measured in volts . the internal resistance , however , may have a more complex origin arising from a series of resistances in the electrodes and electrolyte .
electrical circuit models use equivalent electrical circuits to capture such current - voltage characteristics . in this direction , @xcite consider a hybrid model of kinetic battery and electrical circuit models to fascilitate the derivation of voltage in terms of exponential and polynomial functions of state of charge . building on this approach ,
@xcite fits empirical battery data in order to evaluate battery models for wireless sensor network applications .
the general principle for charge recovery is that of balancing the discharge rate by a positive drift of the available capacity due to the release and transport of stored charges .
such effects should exist as long as the theoretical capacity of the cell has not yet been fully consumed , that is @xmath27 .
the first recovery mechanism to take into account is ( solid - state ) diffusion of charge carriers caused by the build - up of a concentration gradient in the electrolyte during discharge .
the drift of the process is convective flow and a diffusion coefficient controls random variations around the main direction of transport .
diffusion transport of charge carriers might be a slow process which persists even if the load is removed and the battery put to rest , and runs until charge concentrations have reached local equilibrium .
another mechanism for gaining capacity due to recovery is migration of charge - carriers caused by the electric field , as an action of a potential gradient .
the strength of this effect should increase with the gap @xmath28 between maximal and actual capacity .
it appears reasonable to assume that the effect of migration is ongoing whether the battery is under load or at rest .
the final aspect of recovery we wish to include in the modeling scenario is charge transfer , meaning the transfer of charges from electrolyte through an interface to the terminal electrode .
a simplified approach for this effect is that of a friction mechanism , such that a fraction of recovered charges are actually effectuated proportional to the applied load current , either instantaneous current or average current over long time .
the battery models we study are introduced in relation to an arbitrary accumulated discharge function @xmath8 . to engage in a more detailed analysis of battery performance
we consider three stylized examples of @xmath8 , which represent typical discharge patterns for the intended usage of the battery .
_ constant current . _
the first such pattern is that of draining the battery at a constant current @xmath29 which remains the same over the entire battery life until the cell is emptied . clearly , @xmath30 and @xmath31 .
_ deterministic on - off pattern . _
the second example is relevant for the case when we know in advance both the amount of work the battery is supposed to power and the scheduled timing of loads . for such cases
we consider a deterministic pulse - train which consists of a periodic sequence of cycles of equal length .
each cycle begins with an active on - period during which a pulse of constant load is transmitted , followed by an off - period of rest and no load .
specifically we assume that a current @xmath32 [ a ] is drawn continuously during each on - period of length @xmath33 followed by a dormant off - period of length @xmath34 .
hence the cycle duration is @xmath35 and the dutycycle is given by the fraction @xmath36 .
we introduce @xmath37 so that @xmath38 if @xmath4 belongs to an on - period and @xmath39 for @xmath4 in an off - period .
then the consumed capacity is @xmath40 and @xmath41 is a piecewise continuous function with non - decreasing rate @xmath42 .
the average discharge rate equals @xmath43 .
_ random discharge pattern . _
our third stylized example of discharge mechanisms represents the case where no information except average load is available in advance of battery operation .
in this situation we consider completely random discharge with the load to be drawn from the battery per time unit scattered independently and uniformly random in the sense of the poisson process . here
we take @xmath44 , where @xmath45 denotes a standard poisson process on the half line with constant intensity @xmath46 .
this amounts to saying that the battery is drained from energy in small jumps of charge @xmath47 which occur interspaced by independent and exponentially distributed waiting times with expected value @xmath48 .
again the average discharge current is @xmath49 .
the kinetic battery model , originally introduced for lead acid batteries in @xcite , takes the view that remaining capacity of the battery is split in two wells , or compartments , one representing available charge and the other bound charge .
discharge is the consumption of available charge and charge recovery is the flow of matter from the bound well to the available one .
the available capacity @xmath50 in this model is precisely the amount of charge in the available well as function of time .
hence we call @xmath51 , the bound charge and decompose remaining capacity as @xmath52 , @xmath53 , @xmath54 . with the use of the fraction @xmath55 , @xmath56 ,
the two wells are assigned a measure of height given by @xmath57 and @xmath58 .
the principle of the kinetic battery model is that bound charge becomes available at a rate which is proportional to the height difference @xmath59 .
once available , no charges return to the bound state .
thus , @xmath60 dy(t ) & = -k\big(\frac{y(t)}{1-c}-\frac{u(t)}{c}\big)\,dt,\quad & y(0)=t - n , \end{array } \right .
\label{kibameqnsyst}\ ] ] where @xmath61 is a reaction parameter . by assumption , @xmath62 , @xmath63 .
it is convenient therefore to consider the pair @xmath64 . with @xmath65 as an alternative parameter , @xmath66 in ( [ kibameqn ] ) , charge recovery as represented by the factor @xmath67 may be seen as the combination of a migration effect due to the term @xmath28 together with ( negative ) drift @xmath68 . to clarify these connections , ref.@xcite studied a reweighted version of the model . in this paper
we wish to develop these ideas further and hence consider the closely related reweighted model @xmath69 where the parameter @xmath70 controls charge recovery due to migration .
the case @xmath71 reflects migration of additional charge from the bound well adding to the internal charge recovery .
the case @xmath72 represents loss of recovery due to migration .
the linear system ( [ kibameqnplus ] ) is readily solved as @xmath73 for later reference we note that the relevant version of ( [ kibameqnsyst ] ) for this extended case is @xmath74 dy(t ) & = -k_c(cy(t)-(1-c)u(t)+p(n - u(t)))\,dt . \end{array } \right.\ ] ] we emphasize that the discharge profile @xmath5 is arbitrary for this version of the kinetic battery model .
for example , with the random discharge pattern discussed in section [ sec : dischargeprofile ] the integral in ( [ kibameqnplussol ] ) is a stochastic poisson integral and the the corresponding solution is a random process @xmath75 , which may be written @xmath76 where the sum extends over all jumps @xmath77 in @xmath78 $ ] of a poisson process with intensity @xmath79 . of course , the kinetic battery model could only give a crude indication of the processes behind real battery behavior .
in contrast , the diffusion battery model of rakhmatov and vrudhula introduced in @xcite is an approach closer to electrochemical modeling , detailing reaction kinetics and internal transport mechanisms .
both types of models are relevant for engineering battery performance .
a significant step towards unifying these approaches is the extension to spatial versions of the kinetic battery model , where charges move inside of a reservoir of bound charge according to the same local dynamics as the simplest case above .
the diffusion battery model @xcite , views battery operation as the diffusion of ions between two electrodes separated by an electrolyte consisting of a linear region @xmath80 $ ] of the one - dimensional line .
initially , the electroactive species are uniformly distributed over space . over time , the concentration @xmath81 of electroactive species at time @xmath63 and distance @xmath82 from the electrode at @xmath83 develop following fick s laws of diffusion with suitable boundary conditions at both electrode endpoints @xmath83 and @xmath84 . in @xcite
, comparing kinetic and diffusion battery models , jongerden and haverkort introduced the idea of placing a finite number of charge compartments in series along the spatial range and letting charges move between adjacent components according to eq.([kibameqnsyst ] ) .
discharge occurs at the anode which is located in one end point of the spatial interval . starting from a state of fully charged compartments a spatial charge profile develops over time and determines the pace at which the battery is drained .
this system is shown to match the diffusion model when the diffusion equations for @xmath81 are discretized over an equally sized partition of @xmath80 $ ] . in this sense
the diffusion model is a continuous version of the kinetic battery model . in this section
we present further developments of this theory , building on previous work in ref .
indeed , we analyze the effects of charge recovery in the spatial setting and provide in explicit form the available charge and other performance measures in terms of basic parameters of diffusion and migration .
we begin by considering a battery cell consisting of @xmath85 adjacent fluid compartments and a function @xmath86 which gives the charge content in each component over time . here
@xmath87 is the available charge , @xmath88 is a bound well charge for @xmath87 and so on until @xmath89 , which is a bound well charge for @xmath90 . by letting eqn .
( [ kibameqnsyst ] ) act pairwise on adjacent compartments , we obtain the coupled system of linear equations @xmath91 in greater generality , the reweighted version ( [ kibameqnsystplus ] ) yields @xmath92 to see more clearly the structure in this system of equations , we introduce the parameter @xmath93 and employ the notations @xmath94 then @xmath95 here , the parameter @xmath70 controls drift originating from the terms proportional to @xmath28 in ( [ kibameqnsystplus ] ) . in particular , the nominal capacity @xmath2 remains as a parameter in the boundary equations for @xmath87 and @xmath89 .
our next goal is to identify a limiting partial differential equation for the charge concentration profile @xmath50 in the limit @xmath96 as the size of the charge compartments tends to zero and the number of wells goes to infinity .
the case @xmath97 is studied in ref.@xcite and we will use a similar method for the general setting . let @xmath98 and consider the strip @xmath99 partitioned in @xmath85 equal intervals of length @xmath100 . for @xmath101 , @xmath102 , we define @xmath103 and note that @xmath104 and @xmath105 to match spatial and temporal scaling we introduce the scaled parameters @xmath106 , @xmath107 . the relations derived above for @xmath108 ,
@xmath109 imply , for @xmath110 , @xmath111 where @xmath112 is the linear drift operator @xmath113 the additional relations for @xmath87 and @xmath89 correspond to boundary equations for @xmath114 , which attain the form @xmath115 and @xmath116 based on the above relations for the system of @xmath85 compartments one can see that in order to balance all terms in the limit @xmath96 , it is natural to introduce two drift parameters @xmath117 and @xmath118 , both of arbitrary sign , and replace @xmath119 by @xmath120 and @xmath70 by @xmath121 .
then , for large @xmath85 , @xmath122 where @xmath123 is a reaction parameter , @xmath117 a diffusion parameter and @xmath118 a migration parameter .
this gives the approximate system @xmath124 with robin type boundary conditions @xmath125 and @xmath126 we conclude that the relevant limiting equation in the limit @xmath127 , is defined on @xmath128 by @xmath129 here , @xmath130 is the charge density at the boundary of the battery and @xmath131 is the fluid level of a reservoir of bound charge such that @xmath132 is what remains in the reservoir at time @xmath4 . to state a probabilistic representation of the solution to ( [ pdefinal ] ) ,
let @xmath133 denote brownian motion with variance parameter @xmath134 and constant drift @xmath135 .
the parameter @xmath32 in the drift of the brownian motion corresponds to @xmath136 in ( [ pdefinal ] ) .
we assume that @xmath137 is confined to the interval @xmath138 and subject to reflecting boundaries at both end points @xmath139 and @xmath140 .
let @xmath141 be the transition density of @xmath137 so that @xmath142 .
we will use a spectral type representation for @xmath141 known to be @xmath143 the above expression is derived in ref .
@xcite and discussed and compared with an alternative representations in ref .
in particular , for the symmetric case , letting @xmath144 , @xmath145 asymptotically , for @xmath146 as @xmath4 tends to infinity , @xmath147 [ thm : pde ] suppose that the spatial kinetic battery model with reaction parameter @xmath123 , migration parameter @xmath118 , and diffusion parameter @xmath117 is defined on an interval @xmath148 where @xmath2 is the nominal capacity and @xmath1 the theoretical capacity .
let @xmath149 .
a given nonnegative function @xmath150 , is the initial charge profile of the battery and @xmath151 is a given discharge pattern .
we restrict to the range of parameters where the battery model is physically realized , by assuming that @xmath118 , @xmath117 , @xmath2 , @xmath140 , and @xmath152 are such that @xmath153 for @xmath99 , where @xmath154 is defined in ( [ uinftydef ] ) .
then the bound charge profile @xmath155 of the battery , defined as the solution of the pde ( [ pdefinal ] ) , is given for the case @xmath156 by @xmath157 and for the case @xmath158 by @xmath159 where @xmath141 is defined in ( [ transitionprob ] ) .
the special case of diffusion but no migration , which is the pde ( [ pdefinal ] ) with @xmath160 , that is @xmath161 has been studied in ref .
the solution is given by @xmath162 to handle the case of a nonzero migration effect , @xmath163 , in ( [ pdefinal ] ) we first observe that the non - homogeneous term in ( [ eq : pdesolsum ] ) which involves @xmath151 remains the same .
hence it suffices to discuss the solution of of ( [ pdefinal ] ) for the homogeneous case @xmath164 , which represents a battery at rest without discharge current .
we claim that for any drift parameters @xmath117 and @xmath118 , such that @xmath165 , the solution is given by @xmath166 indeed , it is straightforward to verify that for any constant @xmath167 the function @xmath168 satisfies the target equation .
moreover , @xmath169 which shows that boundary condition at @xmath83 is satisfied for @xmath170 .
similarly for the boundary condition at @xmath84 .
finally , to cover the case @xmath158 one shows that @xmath171 equals the sum of the two last terms in ( [ soldelta0 ] ) .
* constant initial charge .
* restricting to the special case of constant initial charge @xmath172 , the solution for @xmath156 is @xmath173 and for @xmath158 , @xmath174 * asymptotic charge profile .
* for a battery at rest , so that @xmath175 , we have by ( [ homogeneousasymptotic ] ) and ( [ pdesolhomogeneous ] ) that @xmath176 is given by @xmath177 in particular , for @xmath178 , @xmath160 yields the truncated exponential function @xmath179 and @xmath180 the trivial asymptotic solution @xmath181 , @xmath182 .
* non - homogeneous case , migration but no diffusion*. the system ( [ pdefinal ] ) for the case @xmath180 and arbitrary @xmath118 , where we also restrict to the initial condition @xmath183 , @xmath184 , takes the form @xmath185 the solution is @xmath186 * varying the discharge profile*. of course , the result in theorem [ thm : pde ] for the capacity reservoir @xmath187 will be obtained in a more or less explicit form depending on which discharge profile @xmath8 applies .
but , in principle , the results of theorem [ thm : pde ] allow comparison of the capacity dynamics under the discharge patterns discussed in section [ sec : dischargeprofile ] , and others .
for example , a family of deterministic on - off patterns with given parameters may be compared to the reference case of constant current discharge .
if the discharge pattern is poisson or otherwise random then the response in capacity and voltage drop will be random as well .
the next subsection is concerned with the case of constant discharge .
the results in this subsection applies to the case of constant current @xmath30 . for simplicity
we assume constant initial capacity @xmath188 .
then the available capacity predicted by the spatial kinetic battery model according to theorem 2.1 is @xmath189 and the corresponding state of charge is @xmath190 .
it is illustrative to consider the available capacity as a function of the remaining capacity stored in the battery .
this point of view is explicit in the simple kinetic battery model under constant current , namely if we consider @xmath64 with @xmath191 and @xmath192 which is the solution of ( [ kibameqnplussol ] ) for the case @xmath30 .
then @xmath193 with @xmath194 given by @xmath195 is an autonomous system , and we may express functionals of @xmath196 , such as voltage @xmath197 , in terms of @xmath198 . in the same spirit we seek to express @xmath199 for the spatial kinetic battery model as a function of remaining capacity . in our model
the relevant remaining capacity function is what is stored in the entire reservoir of size @xmath140 at time @xmath4 , namely @xmath200 by replacing @xmath201 in ( [ nomcapacityspatial ] ) with @xmath202 , @xmath203 which is an autonomous system for the pair @xmath204 .
figure [ fig : phaseplanespatial ] indicates the typical shape of solution curves @xmath193 for fixed , but arbitrary , parameter values @xmath205 ,
@xmath206 , @xmath207 , @xmath208 , and @xmath209 , and for varying @xmath118 and @xmath117 .
the diagonal line from @xmath210 to @xmath211 in the @xmath193 phaseplane represents an idealized situation where all theretical capacity of the battery is consumed at a steady pace .
the additional straight line from @xmath212 to @xmath211 represents a battery which is drained at rate @xmath29 without any recovery of bound charge at all .
in this section we introduce an approach to nonlinear battery modeling using a discrete time markov chain on a set of bivariate states representing available and remaining capacities .
the chain has nonlinear jump probabilities which we arrive at by analyzing a simplified transport system of charges under diffusion and migration .
then we investigate the scaled capacities as the number of slots per time unit tends to infinity and the discharge process converges to that of constant rate . in the limit
we obtain deterministic capacity functions identified as the solution of a nonlinear ordinary differential equation .
informed by these findings we then consider a class of nonlinear odes which can be solved explicitly .
the solutions arising in this manner potentially reflect the nonlinear dynamics of capacity under discharge of the battery .
in addition we indicate two further directions of probabilistic modeling .
one is to study the deviation of the markov chain from its deterministic limit and describe the scaled fluctuations in terms of a diffusion process .
finally we consider the deterministic limit process modified to operate under random discharge , and compare this situation with the previous cases .
we consider a discrete time markov chain @xmath213 defined on the state space @xmath214\times[0,n]$ ] , modeling @xmath215 in time slot $ n$}\\ x_n&=\mbox{available capacity [ ah ] in time slot $ n$. } \end{aligned}\ ] ] we assume that @xmath2 and @xmath1 are integer multiples of @xmath32 and that all jumps are of size @xmath32 .
all jumps in the first coordinate are downwards .
jumps in the second are allowed to be both up and down as long as @xmath216 .
the battery is discharged randomly at constant current @xmath32 with probability @xmath217 per slot . letting @xmath218 where @xmath219 is a sequence of i.i.d random variables with @xmath220
, it follows that @xmath221 is the accumulated discharge at slot @xmath222 and @xmath223 is the remaining charge in the battery after @xmath222 slots .
the expected discharge rate is @xmath224 .
given the sequence @xmath225 as input we model @xmath226 as a markov chain modulated by @xmath227 .
all jumps down of @xmath228 are inherited from the discharge profile and follow those of @xmath227 .
jumps up will occur according to a markovian dynamics chosen so as to reflect the recovery properties of the battery . in slot
@xmath229 the transition probabilities depend on the current state @xmath230 and the current discharge information stored in @xmath231 . in an attempt to model charge recovery the battery cell
is thought to consist of a randomly structured , electroactive material which allows transport of charge carrying species through the electrolyte by liquid or solid state diffusion .
internal charge recovery relies on access to transportation channels of enough connectivity to allow the material to pass from one node to the other .
also , these channels must be `` activated '' by a sufficient amount of previous discharge events . to try to describe such a system ,
we introduce @xmath232 conditional on @xmath233 , the updates @xmath234 and @xmath235 are assumed to have poisson distributions , such that for given nonnegative parameters @xmath236 and @xmath237 @xmath238 and @xmath239 is binomially sampled from @xmath235 , so that @xmath240 in case there is no discharge in slot @xmath222 , that is @xmath241 , then the battery cell is able to recover one unit of charge if both @xmath242 and @xmath243 . hence @xmath244 the dynamics specified by this recursive relation is that , given @xmath245 , if @xmath246 then the transition in slot @xmath229 is @xmath247 and if @xmath248 then @xmath249 together these relations define a bivariate markov chain model @xmath213 with dynamics specified by @xmath250 and , typically , initial condition @xmath251 .
we obtain a drift function and a variance function for the markov chain from @xmath252 and @xmath253 the drift function @xmath254 suggests a relevant , approximating ode for the markov chain . to formalize this limit procedure
it is convenient to introduce a scaling parameter @xmath255 . at scaling level @xmath85
the number of slots per unit time is @xmath85 and the discharge current jump size is @xmath256 rather than @xmath32 .
consider @xmath257 and define for continuous time @xmath63 , @xmath258}.\ ] ] similarly , let @xmath259 be the scaled discharge process with @xmath32 replaced by @xmath256 and put @xmath260 } , \quad v^{(m)}(t)=t-\lambda^{(m)}(t).\ ] ] then @xmath261}z_i = \frac{q\delta[mt]}{m}+\sqrt{\frac{q(1-q)\delta^2}{m } } \frac{1}{\sqrt{m}}\sum_{k=1}^{[mt]}\frac{z_i - q}{\sqrt{q(1-q)}}.\ ] ] by the functional central limit theorem we may introduce a wiener process @xmath262 and for large @xmath85 view @xmath263 as an approximation of the continuous time remaining capacity function @xmath264 , in the sense @xmath265 where @xmath266 and the approximation error is of the order @xmath267 .
moreover , by considering the differential change of @xmath268 over a time interval @xmath269 where @xmath270 , @xmath271 and @xmath272 as above we obtain a deterministic limit equation for large @xmath85 by applying a diffusion approximation with the diffusion term of magnitude @xmath267 . to simplify notation we put @xmath273 then @xmath274 where @xmath275 is another wiener process . since @xmath276 and @xmath277 have simultaneous jumps , @xmath278 and @xmath275 are dependent with a non - zero covariance . as @xmath96 , the stochastic differential equations for @xmath276 and @xmath277 simplify and become the ordinary differential equation @xmath279 & v_t'=-\delta q,&\qquad v_0=t .
\end{array}\ ] ] recalling @xmath280 , the solution is @xmath281 and @xmath282 where @xmath283 hence @xmath284 and so @xmath285 thus , in analogy with the results obtained for the kinetic battery models in the previous section , it follows that @xmath286 is an autonomous system from which we can read off the available capacity as a function of remaining capacity . in the markov chain model of section [ sec : markov ] , we considered random discharge at current @xmath32 with probability @xmath217 per slot , which converged by the law of large numbers to a constant discharge pattern @xmath30 , @xmath280 , under the approximation scheme of section [ sec : approxmarkov ] .
an alternative would be to run the markov chain @xmath233 relative to a given discharge sequence @xmath287 , such that in the scaling limit emerges an on - off discharge pattern as discussed in section [ sec : dischargeprofile ] .
we recall that @xmath288 is the piecewise constant indicator function which is one during periods when the battery is under load and zero otherwise and that @xmath289 is the remaining capacity at time @xmath4 .
as the degree of resolution increases by taking @xmath290 we then expect the process of nominal capacity , @xmath291 , to converge to a deterministic limiting function @xmath292 , which solves the ordinary differential equation @xmath293 then @xmath294 and so , by introducing a generating factor @xmath295 defined by @xmath296 hence @xmath297 we can solve for @xmath298 and obtain @xmath299 with @xmath300 this may be written @xmath301 which is a closed form solution for capacity in terms of the given discharge profile , coded by @xmath288 and @xmath302 with parameters @xmath32 and @xmath217 , and the additional battery parameters @xmath303 and @xmath304 . we have found a deterministic function @xmath305 which satisfies the ordinary differential equation ( [ eq : conttimeapproxode ] ) , in the limit @xmath96 of the scaled markov chain @xmath306 with constant discharge rate @xmath29 .
now we introduce the random quantities @xmath307 in order to study the fluctuations of the scaled markov chain around its deterministic limit @xmath305 . with @xmath278 and @xmath275 as in section [ sec : conttimeapprox ] we obtain for large @xmath85 , @xmath308 where @xmath309 .
it follows by a taylor expansion of @xmath310 that @xmath311 in the scaling limit @xmath96 we therefore expect that the fluctuation process @xmath312 converges to a diffusion process @xmath313 , such that @xmath314 this stochastic differential equation for @xmath313 can be seen as a generalized ornstein uhlenbeck process @xmath315 with time - inhomogeneous drift and variance coefficients @xmath316 and @xmath317 , which is also modulated by an additional , random , drift @xmath318 . to derive the capacity dynamics in ( [ markovode ] ) we were guided by a markov chain argument based on a simplified view of charge transport inside the battery cell . in this final section
we mention a more general class of deterministic non - linear recovery models for the dynamics of nominal capacity and state of charge .
again we write @xmath319 for nominal capacity and @xmath320 for the remaining capacity in the cell as functions of time @xmath63 . here
@xmath321 and @xmath5 is the discharge process with average discharge current @xmath31 .
the general principle for charge recovery is that of balancing the discharge rate in @xmath319 by a positive drift of the nominal capacity due to the release and transport of stored charges .
it is reasonable that such effects are proportional to the applied average load @xmath322 and exist as long as the theoretical capacity of the cell has not yet been fully consumed , that is @xmath323 .
furthermore , the gain in capacity due to recovery depends on the migration of charge - carriers , and the strength of this effect should increase with the gap @xmath324 between maximal and actual capacity . hence , to capture charge recovery in the framework of one - dimensional differential equations ,
we may let @xmath325 $ ] and @xmath326 $ ] be non - decreasing functions with @xmath327 , @xmath328 , and consider equations of the generic shape @xmath329 where @xmath330 is battery life defined as the maximal time @xmath4 for which @xmath331 .
the special case of ( [ genericode ] ) where @xmath332 and @xmath333 is a positive constant which measures the strength of migration of charges due to the existing electric field in the battery , is given by @xmath334 because of the separation of variables we may write @xmath335 for example , taking the simplest case @xmath30 , @xmath336 equivalently , @xmath337 it is immediate in this model that the pair @xmath305 is autonomous so that available capacity @xmath338 can be viewed as a function of remaining capacity only . moreover , the capacity dynamics is load - invariant in the sense that the curve @xmath339 does not depend on @xmath29 .
in other words , batteries drained using different choices of @xmath29 will exhibit different battery life , longer the smaller intensity of the current , but the used capacity @xmath340 at end of life will be the same in each case . in the framework of the nonlinear ode approach we have obtained , just as for the linear models studied previously , a phase plane relation @xmath341 for the capacity dynamics .
via state - of - charge @xmath342 we may proceed as before to modeling the corresponding voltage . without going into details
this will give us some capacity threshold @xmath343 below which the battery is no more functioning .
then the unused capacity that remains in the battery at the end of its life time is the unique solution @xmath344 of @xmath345 .
thus , the delivered capacity is @xmath346 and the gained capacity is @xmath347 .
our model allows for some qualitative conclusions about these quantities as well as numerical studies of special cases .
various discharge processes @xmath8 can be compared against experimental data and relevant parameters estimated . to give an indication of this type of work , figure [ fig : dischargeprofile ]
shows the result of repeated independent simulations based on ( [ expode ] ) with @xmath348 and @xmath8 a random poisson discharge process as described in section [ sec : dischargeprofile ] .
the parameters used for these particular simulations are chosen arbitrarily such that the output appears to mimic that of a real battery .
the upper panel shows the phase plane traces of the resulting solutions @xmath349 .
the lower panel shows the corresponding random paths of the state - of - charge @xmath350 as function of time .
also in the lower panel we have superimposed ( solid red curve ) the state - of - charge for the case of constant discharge with the same average load , now given by the explicit solution of equation ( [ expode ] ) with @xmath30 .
i. kaj , v. konane , analytical and stochastic modelling of battery cell dynamics .
_ proceed . 19th intern .
analytic and stochastic modelling techn . and appl . _ k. al - begain , d. fiems , and j .-
vincent ( eds . ) : asmta 2012 , lecture notes in computer science , vol 7314 ( 2012 ) , 240254 .
v. ramadesigan , p.w.c northrop , s. de , s. sananthagopalan , r.d .
braatz and v.r .
subramanian , modeling and simulation of lithium - ion batteries from a systems engineering perspective .
j. electrochem .
* 159*:3 ( 2012 ) , r31-r45 . c. rohner , l.m .
feeney , and p. gunningberg , evaluating battery models in wireless sensor networks .
11th international conference on wired / wireless internet communication ( wwic ) , lecture notes in computer science vol 7889 , 2013 . v. subramaniam , v. boovaragavan , v. ramadesigan , m. arabandi , mathematical model reformulation for lithium - ion battery simulations : galvanostatic boundary conditions . j. electrochem .
soc . * 156*:4 ( 2009 ) , a260-a271 .
svensson , the term structure of interest rate differentials in a target zone , theory and swedish data .
seminar paper 466 , institute for international economic studies , stockholm university 1990 . | in this paper we review several approaches to mathematical modeling of simple battery cells and develop these ideas further with emphasis on charge recovery and the response behavior of batteries to given external load .
we focus on models which use few parameters and basic battery data , rather than detailed reaction and material characteristics of a specific battery cell chemistry , starting with the coupled ode linear dynamics of the kinetic battery model .
we show that a related system of pde with robin type boundary conditions arises in the limiting regime of a spatial kinetic battery model , and provide a new probabilistic representation of the solution in terms of brownian motion with drift reflected at the boundaries on both sides of a finite interval . to compare linear and nonlinear dynamics in kinetic and stochastic battery models we study markov chains with states representing available and remaining capacities of the battery .
a natural scaling limit leads to a class of nonlinear ode , which can be solved explicitly and compared with the capacities obtained for the linear models . to indicate the potential use of the modeling we discuss briefly comparison of discharge profiles and effects on battery performance .
battery lifetime ; state - of - charge ; charge recovery ; probabilistic solution of pde ; robin boundary condition ; nonlinear ode s h |
scalar field , associated with spin-@xmath0 particles in quantum field theory , has been an object of great interest for physicists in the latest years .
one of the main reasons is that the models with scalar fields are relatively simple , which allows us to probe the detailed features of the more complicated physical system . in cosmology ,
scalar fields can be considered as candidate to explain the inflation of the early universe @xcite and the accelerated expansion of the current universe @xcite . in the standard model of particle physics
, the scalar field presents as the higgs boson @xcite , which would help to explain the origin of mass in the universe .
moreover , it has been found that scalar field plays the important roles in other fundamental physical theories , such as , jordan - brans - dicke theory @xcite , kaluza - klein compactification theory @xcite and superstring theory @xcite , and so on .
in general , the action contained scalar fields in einstein s theory of gravity is @xmath1+s_m,\label{act1}\end{aligned}\ ] ] where @xmath2 , @xmath3 and @xmath4 are corresponding to scalar field , ricci scalar and scalar potential , respectively .
the term @xmath5 represents the coupling between ricci scalar @xmath3 and the scalar field @xmath2 .
the dynamical behavior of the scalar field in the theory ( [ act1 ] ) have been investigated very extensively in the modern physics including cosmology and black hole physics .
the more general form of the action contained scalar field in other theories of gravity is @xmath6+s_m,\label{act2}\end{aligned}\ ] ] here @xmath7 and @xmath8 are arbitrary functions of the corresponding variables .
obviously , the more coupling between scalar field and curvature are considered in these extensive theories .
the non - minimal coupling between the derivative of scalar field and the spacetime curvature may appear firstly in some kaluza - klein theories @xcite .
amendola @xcite considered the most general theory of gravity with the lagrangian linear in the ricci scalar , quadratic in @xmath2 , in which the coupling terms have the forms as follows @xmath9 and then he studied the dynamical evolution of the scalar field in the cosmology by considering only the derivative coupling term @xmath10 and obtained some analytical inflationary solutions @xcite .
@xcite investigated a more general model of containing coupling terms @xmath11 and @xmath10 , and found that the de sitter spacetime is an attractor solution in the model .
recently , daniel and caldwell @xcite obtained the constraints on the theory with the derivative coupling term of @xmath10 by solar system tests .
in general , a theory with derivative couplings could lead to that both the einstein equations and the equation of motion for the scalar are the fourth - order differential equations .
however , sushkov @xcite studied recently the model in which the kinetic term of the scalar field only coupled with the einstein tensor and found that the equation of motion for the scalar field can be reduced to second - order differential equation .
this means that the theory is a good " dynamical theory from the point of view of physics .
gao @xcite investigated the cosmic evolution of a scalar field with the kinetic term coupling to more than one einstein tensors and found the scalar field presents some very interesting characters .
he found that the scalar field behaves exactly as the pressureless matter if the kinetic term is coupled to one einstein tensor and acts nearly as a dynamic cosmological constant if it couples with more than one einstein tensors .
the similar investigations have been considered in refs.@xcite .
these results will excite more efforts to be focused on the study of the scalar field coupled with tensors in the more general cases .
since black hole is another fascinating object in modern physics , it is of interest to extend the study of the properties of the scalar field when it is kinetically coupled to the einstein tensors in the background of a black hole . in this letter
, we will investigate the greybody factor and hawking radiation of the scalar field coupling only to the einstein tensor @xmath12 in the reissner - nordstrm black hole spacetime by using the matching technique , which has been widely used in evaluating the absorption probabilities and hawking radiations of various black holes @xcite .
we find that the presence of the coupling terms enhances both the absorption probability and hawking radiation of the scalar field in the black hole spacetime .
moreover , we also find that for the stronger coupling , the absorption probability and hawking radiation of the scalar field increase with the charge of the black hole , which is different from those of scalar one without coupling to einstein s tensor .
the letter is organized as follows : in the following section we will introduce the action of a scalar field coupling to einstein s tensor and derive its master equation in the reissner - nordstrm black hole spacetime . in sec.iii
, we obtain the expression of the absorption probability in the low - energy limit by using the matching technique . in section iv
, we will calculate the absorption probability and the luminosity of hawking radiation for the coupled scalar field .
finally , in the last section we will include our conclusions .
let us consider the action of the scalar field coupling to the einstein s tensor @xmath12 in the curved spacetime @xcite , @xmath13.\label{acts}\end{aligned}\ ] ] the coupling between einstein s tensor @xmath12 and the scalar field @xmath2 is represented by the term @xmath14 , where @xmath15 is coupling constant with dimensions of length - squared . in general , the presence of such a coupling term brings some effects to the original metric of the background .
however , it is very difficult for us to obtain an analytic solution for the action ( [ acts ] ) . actually , comparing with the mass of the black hole
, one can find that the energy of a scalar field is very tiny so that its back - reaction effects on the background can be neglected exactly . here , we treat the weak external scalar filed as a probe field , and then study the effects of the coupling constant @xmath15 on the greybody factor and hawking radiation of the scalar filed in the background of a black hole spacetime . varying the action with respect to @xmath2 ,
one can obtain the modified klein - gordon equation @xmath16 = 0,\label{we}\end{aligned}\ ] ] which is a second order differential equation .
obviously , all the components of the tensor @xmath12 vanish in the schwarzschild black hole spacetime because it is the vacuum solution of the einstein s field equation .
thus , we can not probe the effect of the coupling term on the greybody factor and hawking radiation in the background of a schwarzschild black hole .
the simplest black hole with the non - zero components of the tensor @xmath12 is reissner - nordstrm one , which can be described by the metric as follows @xmath17 with @xmath18 where @xmath19 is the mass and @xmath20 is the charge of the black hole .
the einstein s tensor @xmath12 for the metric ( [ m1 ] ) has a form @xmath21 adopting to the spherical harmonics @xmath22 , we find that the equation ( [ we ] ) can be separable and the radial function @xmath23 obeys to @xmath24 + \bigg[\bigg(1+\frac{\eta q^2}{r^4}\bigg)\frac{\omega^2}{f}-\bigg(1-\frac{\eta q^2}{r^4}\bigg)\frac{e_{lm}}{r^2}\bigg]r(r)=0,\label{radial}\end{aligned}\ ] ] where @xmath25 is the eigenvalues coming from the angular equation .
clearly , the radial equation ( [ radial ] ) contains the coupling constant @xmath15 , which means that the presence of the coupling term will change the evolution of the scalar field in the reissner - nordstrm black hole spacetime .
moveover , as the charge @xmath20 vanishes , it is easy to obtain that the effects of the coupling constant @xmath15 disappears and the radial equation ( [ radial ] ) reduces to that in the schwarzschild black hole spacetime . the solution of the radial function @xmath23 will help us to obtain the absorption probability @xmath26 and the luminosity of hawking radiation for a scalar field coupling to einstein s tensor in the reissner - nordstrm black hole spacetime .
in this section , we will present an analytic expressions for the greybody factors for the emission of a scalar field coupling to einstein s tensor in the reissner - nordstrm black hole spacetime .
since the radial equation ( [ radial ] ) is generally nonlinear , it is very difficult to obtain its analytic solution .
however , following refs.@xcite , we can provide an approximated solution of the radial equation ( [ radial ] ) by employing the matching technique .
firstly , we can derive the analytic solutions in the near horizon ( @xmath27 ) and far - field @xmath28 regimes in the low - energy limit ( i.e. , @xmath29 and @xmath30 ) , respectively .
then , we can match smoothly these two radial functions @xmath23 in an intermediate region since the wave function in physics is continuous everywhere . through this matching technique
we can construct a smooth analytical solution of the radial equation valid throughout the entire spacetime and extract further the ratio between two coefficients @xmath31 and @xmath32 in eq .
( [ rf6 ] ) , which helps us define the greybody factor .
the main reason that we here adopt to the low - energy approximation is that in this limit we can neglect the back - reaction of a scalar field to the background metric during the emission process .
moreover , it is well known that the greybody factor modifies the spectrum of emitted particles from that of a perfect thermal blackbody . however , in the high - energy regime , the greybody factor is independent of the energy @xmath33 of the particle and the spectrum is exactly like that of a perfect blackbody for every particle species @xcite , while in the low - energy regime , it encodes information about the near horizon structure of a black hole and about the particles emitted by the black hole .
now , we focus on the near - horizon regime and perform the following transformation of the radial variable as in refs .
@xcite @xmath34 with @xmath35 the equation ( [ radial ] ) near the horizon @xmath36 can be rewritten as @xmath37r(f)=0,\label{r1}\end{aligned}\ ] ] where @xmath38 making the field redefinition @xmath39 , one can find that the equation ( [ r1 ] ) can be rewritten as a form of the hypergeometric equation @xmath40\frac{d f(f)}{d f}-ab f(f)=0,\label{r2}\end{aligned}\ ] ] with @xmath41 considering the constraint coming from coefficient of @xmath42 , one can easy to obtain that the power coefficients @xmath43 and @xmath44 satisfy @xmath45 and @xmath46=0,\end{aligned}\ ] ] respectively .
these two equations admit that the parameters @xmath43 and @xmath44 have the forms @xmath47 } \;\bigg\},\label{bet}\end{aligned}\ ] ] following the operation in ref.@xcite and using the boundary condition that no outgoing mode exists near the horizon , we can obtain that the parameters @xmath48 and @xmath49 .
thus the asymptotic solution near horizon has the form @xmath50 where @xmath51 is an arbitrary constant .
let us now to stretch smoothly the near horizon solution to the intermediate zone . as done in ref.@xcite ,
we can make use of the property of the hypergeometric function @xcite and change its argument in the near horizon solution from @xmath7 to @xmath52 @xmath53.\label{r2}\end{aligned}\ ] ] as @xmath54 , the function @xmath55 can be approximated as @xmath56 and then the near horizon solution ( [ r2 ] ) can be simplified further to @xmath57 with @xmath58 @xmath59 in order to obtain a solution in the far field region , we expand the wave equation ( [ radial ] ) as a power series in @xmath60 and keep only the leading terms @xmath61 this is usual bessel equation .
thus the solution of radial master equation ( [ radial ] ) in the far - field limit can be expressed as @xmath62,\label{rf}\end{aligned}\ ] ] where @xmath63 and @xmath64 are the first and second kind bessel functions , @xmath65 .
@xmath66 and @xmath67 are integration constants . in order to stretch the far - field solution ( [ rf ] ) towards small radial coordinate ,
we take the limit @xmath68 and obtain @xmath69 in the low - energy limit , the two power coefficients in eq.([rn2 ] ) can be approximated as @xmath70 respectively . by matching the corresponding coefficients between eqs.([rn2 ] ) and ( [ rfn2 ] ) , we can obtain two relations between @xmath71 and @xmath72 .
removing @xmath73 , we can obtain the ratio between the coefficients @xmath74 @xmath75^{2l+1 } \bigg(\frac{2l+1}{2}\bigg)\gamma^2(l+\frac{1}{2 } ) \nonumber\\ & \times&\;\frac { \gamma(c - a - b)\gamma(a)\gamma(b)}{\gamma(a+b - c)\gamma(c - a)\gamma(c - b)}. \label{bb}\end{aligned}\ ] ] in the asymptotic region @xmath76 , the solution in the far - field can be expressed as @xmath77 the absorption probability can be calculated by @xmath78 inserting the expression of @xmath79 ( [ bb ] ) into eq.([gfa ] ) , we can probe the properties of absorption probability for the scalar field coupled with einstein s tensor in the charged black hole spacetime in the low - energy limit .
we are now in a position to calculate the absorption probability and discuss hawking radiation of a scalar field coupling to einstein s tensor in the background of a reissner - nordstrm black hole . in fig.1 , we fixed the coupling constant @xmath15 and plotted the change of the absorption probability of a scalar particle with the charge @xmath20 for the first partial waves ( @xmath80 ) in the reissner - nordstrm black hole .
one can easily see that with for the smaller @xmath15 the absorption probability @xmath81 decreases with the charge @xmath20 of black hole , which is similar to that for the usual scalar field without coupling to einstein s tensor . however , for the larger @xmath15 , the absorption probability @xmath81 increases as the charge @xmath20 increase , which means that the stronger coupling between the scalar field and einstein s tensor changes the properties of the absorption probability of scalar field in the black hole spacetime . this phenomenon has not been observed elsewhere . from fig.2
, we also find that the absorption probability increases with the increase of the coupling constant @xmath15 for fixed values of charge @xmath82 .
these results about the absorption probability hold true for other values of @xmath83 .
it is also shown in figs . 3 and 4 , in which we plotted the dependence of the absorption probability on the angular index .
moreover , we see the suppression of @xmath84 as the values of the angular index increase . this can be explained by a fact that the absorption probability @xmath85 in the low - energy approximation @xmath86 @xcite , which means that the first partial wave dominates over all others in the absorption probability .
it is similar to that of the scalar field without coupling to einstein s tensor as shown in refs.@xcite .
[ fig1 ] of a scalar field with the charge @xmath20 in the reissner - nordstrm black hole for fixed @xmath80 .
the coupling constant @xmath15 is set by @xmath87 in the left and by @xmath88 in the right .
the solid , dashed , dash - dotted and dotted lines are corresponding to the cases with @xmath89 , respectively .
we set @xmath90.,title="fig:",width=302 ] of a scalar field with the charge @xmath20 in the reissner - nordstrm black hole for fixed @xmath80 .
the coupling constant @xmath15 is set by @xmath87 in the left and by @xmath88 in the right .
the solid , dashed , dash - dotted and dotted lines are corresponding to the cases with @xmath89 , respectively .
we set @xmath90.,title="fig:",width=302 ] [ fig2 ] of a scalar field on the coupling constant @xmath15 in the reissner - nordstrm black hole for fixed @xmath80 and @xmath82 .
the solid , dashed , dash - dotted and dotted lines are corresponding to the cases with @xmath91 , respectively .
we set @xmath90.,width=302 ] [ fig3 ] of a scalar field with the charge @xmath20 in the reissner - nordstrm black hole for fixed @xmath92 .
the coupling constant @xmath15 is set by @xmath87 in the left and by @xmath88 in the right .
the solid , dashed , dash - dotted and dotted lines are corresponding to the cases with @xmath89 , respectively .
we set @xmath90.,title="fig:",width=302 ] of a scalar field with the charge @xmath20 in the reissner - nordstrm black hole for fixed @xmath92 .
the coupling constant @xmath15 is set by @xmath87 in the left and by @xmath88 in the right .
the solid , dashed , dash - dotted and dotted lines are corresponding to the cases with @xmath89 , respectively .
we set @xmath90.,title="fig:",width=302 ] [ fig4 ] of a scalar field on the coupling constant @xmath15 in the reissner - nordstrm black hole for fixed @xmath92 and @xmath82 .
the solid , dashed , dash - dotted and dotted lines are corresponding to the cases with @xmath91 , respectively .
we set @xmath90,width=302 ] now let us turn to study the luminosity of the hawking radiation of a scalar field coupling to einstein s tensor in the background of a reissner - nordstrm black hole . from above discussion , we know that in the low - energy approximation @xmath86 the mode @xmath80 plays a dominant role in the greybody factor , which means that in the scalar emission the major contribution comes from the mode @xmath80 since the power emission spectra of a scalar particle is in direct proportion to @xmath26 . thus , we here consider only the zero mode ( @xmath80 ) and study the effect of the coupling constant @xmath15 on the luminosity of the hawking radiation in the background spacetime . performing an analysis similar to that in @xcite , we can obtain that the greybody factor ( [ gfa ] ) in the low - energy limit has a form @xmath93 combining it with hawking temperature @xmath94 of reissner - nordstrm black hole , the luminosity of the hawking radiation for the scalar field with coupling to einstein s tensor is given by latexmath:[\ ] ] this means that @xmath105 increases with the increase of @xmath15 , which leads to the stronger hawking radiation .
in this letter , we have studied the greybody factor and hawking radiation for a scalar field coupling to einstein s tensor in the background of reissner - nordstrm black hole in the low - energy approximation .
we have found that the absorption probability and hawking radiation depend on the coupling between the scalar field and einstein s tensor .
the presence of the coupling enhances both the absorption probability and hawking radiation of the scalar field in the black hole spacetime .
moreover , for the weaker coupling , we also find that the absorption probability and hawking radiation decreases with the charge @xmath20 of the black hole .
it is similar to that of the scalar field without coupling to einstein s tensor .
however , for the stronger coupling , the charge @xmath20 enhances the absorption probability and hawking radiation of the black hole , which could provide a way to detect whether there exist a coupling between the scalar field and einstein s tensor or not .
it would be of interest to generalize our study to other black hole spacetimes , such as rotating black holes etc .
work in this direction will be reported in the future .
this work was partially supported by the national natural science foundation of china under grant no.10875041 , the program for changjiang scholars and innovative research team in university ( pcsirt , no .
irt0964 ) and the construct program of key disciplines in hunan province .
j. jing s work was partially supported by the national natural science foundation of china under grant no.10675045 , no.10875040 and
no.10935013 ; 973 program grant no .
2010cb833004 and the hunan provincial natural science foundation of china under grant no.08jj3010 .
b. ratra and j. peebles , phys .
d * 37 * ( 1988 ) 3406 ; c. wetterich , nucl .
b * 302 * ( 1988 ) 668 ; r. r. caldwell , r. dave and p. j. steinhardt , phys .
. lett . * 80 * ( 1988 ) 1582 ; m. doran and j. jaeckel , phys .
d * 66 * ( 2002 ) 043519 .
r. r. caldwell , phys .
b * 545 * ( 2002 ) 23 ; b. mcinnes , j. high energy phys . * 08 * ( 2002 ) 029 ; s. nojiri and s. d. odintsov , phys .
b * 562 * ( 2003 ) 147 ; l. p. chimento and r. lazkoz , phys . rev
* 91 * ( 2003 ) 211301 ; b. boisseau , g. esposito - farese , d. polarski , alexei a. starobinsky , phys .
* 85 * ( 2000 ) 2236 ; r. gannouji , d. polarski , a. ranquet , a. a. starobinskyjcap * 0609 * ( 2006 ) 016 . c. m. harris and p. kanti , jhep * 0310 * ( 2003 ) 014 ; p. kanti , int . j. mod
. phys . a * 19 * ( 2004 ) 4899 ; p. argyres , s. dimopoulos and j. march - russell , phys .
lett . b*441 * ( 1998 ) 96 ; t. banks and w. fischler , hep - th/9906038 ; r. emparan , g. t. horowitz and r. c. myers , phys .
* 85 * ( 2000 ) 499 .
e. jung and d. k. park , nucl .
b * 717 * ( 2005 ) 272 ; n. sanchez , phys . rev .
d * 18 * ( 1978 ) 1030 ; e. jung and d. k. park , class . quantum grav .
* 21 * ( 2004 ) 3717 ; e. jung , s. h. kim and d. k. park , jhep * 0409 * ( 2004 ) 005 .
a. s. majumdar , n. mukherjee , int .
d * 14 * ( 2005 ) 1095 ; g. kofinas , e. papantonopoulos and v. zamarias , phys .
d * 66 * ( 2002 ) 104028 ; g. kofinas , e. papantonopoulos and v. zamarias , astrophys .
space sci .
* 283 * ( 2003 ) 685 ; a. n. aliev , a. e. gumrukcuoglu , phys .
d * 71 * ( 2005 ) 104027 ; s. kar , s. majumdar , int . j. mod
. phys . a * 21 * ( 2006 ) 6087 ; s. kar , s. majumdar , phys . rev .
d * 74 * ( 2006 ) 066003 ; s. kar , phys . rev .
d * 74 * ( 2006 ) 126002 .
e. jung , s. h. kim and d. k. park , phys .
b * 615 * ( 2005 ) 273 ; e. jung , s. h. kim and d. k. park , phys .
b * 619 * ( 2005 ) 347 ; d. ida , k. oda and s. c. park , phys .
d * 67 * ( 2003 ) 064025 ; g. duffy , c. harris , p. kanti and e. winstanley , jhep * 0509 * ( 2005 ) 041 ; m. casals , p. kanti and e. winstanley , jhep * 0602 * ( 2006 ) 051 ; e. jung and d. k. park , nucl .
b * 731 * ( 2005 ) 171 ; a. s. cornell , w. naylor and m. sasaki , jhep * 0602 * ( 2006 ) 012 ; v. p. frolov , d. stojkovic , phys .
* 89 * ( 2002 ) 151302 ; valeri p. frolov , dejan stojkovic , phys .
d * 66 * ( 2002 ) 084002 ; d. stojkovic , phys .
* 94 * ( 2005 ) 011603 . | = 0.6 cm * abstract * we study the greybody factor and hawking radiation for a scalar field coupling to einstein s tensor in the background of reissner - nordstrm black hole in the low - energy approximation .
we find that the presence of the coupling terms modifies the standard results in the greybody factor and hawking radiation .
our results show that both the absorption probability and hawking radiation increase with the coupling constant .
moreover , we also find that for the stronger coupling , the charge of black hole enhances the absorption probability and hawking radiation of the scalar field , which is different from those of ones without coupling to einstein s tensor in the black hole spacetime .
= 0.8 cm |
the exploration of phase transitions in simple , one component nonequilibrium models has attracted considerable interest @xcite , and important steps towards identifying the related universality classes have been achieved @xcite . in nonequilibrium models phase transitions may occur even in one - dimensional systems : the well known arguments , due to landau and to van hove @xcite , against phase transitions in one - dimensional systems with short - range interactions , do not apply in the absence of detailed balance . however in low dimensions the effect of fluctuations is stronger , making continuous phase transitions more common .
( a familiar example is the three - state potts model , which exhibits a continuous transition in two dimensions , and a discontinuous one for @xmath3 . ) in one dimension , discontinuous phase transitions have been found in models with long - range interactions @xcite , or a conserved density @xcite , and in multi - component systems @xcite .
compact directed percolation ( cdp ) has a discontinuous transition between a pair of absorbing states ( all sites full or all empty ) @xcite ; a similar transition between absorbing states is found in the one - dimensional ziff - gulari - barshad model @xcite .
a discontinuous transition between and active phase and an absorbing one in a single - component model was claimed for the triplet creation model ( tcm ) @xcite , which does not possess a conservation law or long - range interactions .
this model features particle reactions @xmath0 , @xmath1 and explicit diffusion ( hopping ) @xcite . on increasing the diffusion probability , a crossover from a continuous to a discontinuous phase transition
was detected in simulations and cluster mean - field approximations .
similar behavior was also reported in a stochastic cellular automaton @xcite .
subsequently , hinrichsen argued that in one dimension , discontinuous transitions between an active and an absorbing state can not exist in models like the tcm @xcite .
the original findings for the tcm @xcite were nevertheless confirmed in spreading simulations by cardoso and fontanari @xcite and in fixed order - parameter simulations by fiore and de oliveira @xcite .
the spreading exponents are shown in @xcite to be those of compact directed percolation ( cdp ) @xcite ; a tricritical point is suggested for a diffusion probability @xmath4 .
very recently park @xcite reported simulation results that again support a continuous phase transition , belonging to the directed percolation ( dp ) universality class , at high diffusion rates . recently ,
a field theoretic analysis of bosonic reaction - diffusion ( rd ) models led to a hypothesis @xcite , based on a general phase transition classification scheme : _ bosonic , one - component rd systems with @xmath2-particle creation and @xmath5-particle annihilation always exhibit a first - order transition if @xmath6_. this is indeed the case above the upper critical dimension ( see @xcite ) .
however in bosonic models one has to introduce a higher - order coagulation term @xmath7 with ( @xmath8 ) , to avoid an infinite particle density in the active phase .
furthermore the topological phase space method used in @xcite deals with the reactions ( creation and annihilation ) , but does not take into account the effect of diffusion , which turns out to be relevant in some cases , when different reactions compete @xcite . in this work we study the tcm in an effort to determine whether multi - particle creation , combined with rapid diffusion , can overcome fluctuations and generate a discontinuous phase transition in one dimension .
this is a problem of longstanding interest in nonequilibrium statistical physics , and is related to the existence of first - order depinning transition in nonequilibrium wetting ( i.e. , in a system with multiplicative noise , with an attractive wall ) @xcite .
the remainder of this paper is organized as follows . in sec .
ii we define the model , and review applicable simulation methods and previous results regarding the nature of the phase transition .
section iii is devoted to a discussion of @xmath2-site approximations , and sec .
iv to our simulation results . in sec .
v we use these results to motivate a simplified description of the model in the high diffusion rate regime , and discuss the nature of the transition using this mapping . finally in sec .
vi we summarize our findings .
the tcm is defined on a lattice , with each site either vacant or occupied by a particle ; multiple occupancy is forbidden @xcite . in the one - dimensional tcm , a particle ( @xmath9 ) attempts diffusion at rate @xmath10 , creation ( @xmath11 ) at rate @xmath12 , and is annihilated ( @xmath13 ) at rate @xmath14 . in a diffusion attempt ,
one of the nearest neighbor ( nn ) sites of the particle is chosen at random , and the particle jumps to this site if it is empty .
if the target site is occupied , the configuration remains the same . in a creation attempt ,
if both nn sites of the particle are occupied , then one of the second - neighbor sites of the central particle is chosen at random , and if this site is empty , a new particle is placed there .
if the conditions of two occupied nn sites and an empty target site are not fulfilled , the configuration does not change .
annihilation occurs independently of the states of neighboring sites .
the configuration with all sites empty is absorbing . since the sum of these transition rates is unity , the total transition rate in a system with @xmath15 particles is simply @xmath15 . in simulations ,
the time increment associated with each attempted event ( whether accepted or not ) is @xmath16 , and one monte carlo step ( mcs ) corresponds to an accumulated time increment of unity . in ref .
@xcite the one - dimensional tcm was shown to exhibit a phase transition between the active and absorbing states ; the transition was found to be continuous ( and in the dp universality class ) for smaller diffusion rates , but discontinuous for large @xmath17 . by a discontinuous transition
we mean one in which the order parameter is a discontinuous function of the relevant control parameter(s ) , in the infinite - size limit . in the tcm the order parameter is the particle density @xmath18 , and the control parameters are @xmath19 and @xmath17 . since one of the phases is absorbing , at a discontinuous transition @xmath18 should jump between zero and a finite value .
the characterization of a transition as continuous or discontinuous in numerical simulations is fraught with difficulties : finite - size rounding can mask the discontinuity , and any finite system must eventually become trapped in the absorbing state . to circumvent these problems , a number of strategies have been proposed .
_ hysteresis with a weak source . _
a characteristic feature of discontinuous phase transitions is hysteresis .
if one of the phases is absorbing , however , hysteresis can not be observed simply by varying a control parameter , since the absorbing phase allows no escape .
bideaux , boccara and chat @xcite showed that when the transition is discontinuous , adding a weak source of activity changes the absorbing and active phases to _ low - activity _ and _ high - activity _ phases , respectively .
one may then observe a hysteresis loop between these phases , on varying the control parameter .
this approach was used in @xcite to demonstrate a discontinuous phase transition in a probabilistic cellular automaton , and was applied to the tcm in @xcite , yielding a hysteresis loop .
below , we shall revisit the question of scaling under a weak source .
_ conserved order parameter simulations . _
in conserved order parameter simulations @xcite , particles are neither created nor destroyed .
changes in configuration occur through particle jumps , which can be of any size up to that of the entire system , in a manner that respects the local rules of the process .
the simulation yields an estimate for the control parameter value corresponding to the chosen order parameter density .
using this method , fiore and de oliveira found evidence for a discontinuous transition in both the tcm and the related pair creation model ( with creation reaction @xmath20 ) at high diffusion rates @xcite . _ quasistationary ( qs ) simulation .
_ as in conserved order parameter simulations , qs simulation removes the absorbing state from the dynamics , but in a manner that samples the quasistationary probability distribution ( i.e. , conditioned on survival ) @xcite .
a study of the tcm using this method @xcite showed that as the system size tends to infinity , the qs order parameter appears to develop a discontinuity between zero and a positive value , as @xmath19 is varied at a high diffusion rate , @xmath21 .
( for a finite system the discontinuity is of course rounded . )
a study of the tcm with biased diffusion ( hopping in one direction only ) yielded evidence of a sharp discontinuity @xcite .
_ spreading simulations .
_ studies of the spread of activity , starting from a seed at the origin , have long been employed to characterize continuous phase transitions to an absorbing state @xcite . at the critical point ,
the survival probability @xmath22 , mean number of active sites @xmath23 , and mean - square distance @xmath24 of active sites from the origin , all follow power laws . at a discontinuous transition ,
there is in principle no reason to expect scale - invariant spreading dynamics .
nevertheless , in the case of the tcm , cardoso and fontanari @xcite demonstrated power - law spreading at the transition point , @xmath25 , for @xmath21 .
the scaling exponents were identified as those of cdp , which , as noted above , suffers a discontinuous transition between a pair of symmetric absorbing states . _ interface motion . _
suppose we prepare the system with all sites occupied , allow it to relax to the qs state , and then remove all particles from half of the lattice .
in the subsequent evolution , the interface between active and inactive regions broadens due to diffusion , and in general drifts toward one region or the other .
below we report studies showing that the drift velocity is proportional to @xmath26 . in a related analysis ,
we initialize the system with all sites in the region @xmath27 occupied , and sites @xmath28 empty , and study the long - time survival probability @xmath29 . at the transition ,
the dependence of @xmath29 on @xmath30 is consistent with independent , randomly diffusing interfaces , as in cdp .
this result supports the existence of two phases , one absorbing , the other active , separated by a large gap in density .
the two phases do not coexist : the fluctuating interfaces eventually meet , and one of the phases is lost from the system .
summarizing , the above mentioned studies , some from the recent literature , others to be reported below , provide evidence for a discontinuity in the qs order parameter , for hysteresis , and for a connection between the tcm at high diffusion rate and compact directed percolation .
some years ago , hinrichsen presented an argument to the effect that discontinuous phase transitions between an active and an absorbing state are impossible in one - dimensional systems with local interactions , and without additional conservation laws , special boundary conditions , or macroscopic currents @xcite .
the argument is based on the observation that the effective surface tension of interfaces in such systems does not depend on the size of the domains they delimit .
hinrichsen s argument prohibits the presence of fixed , stable boundaries between coexisting phases ; as noted , no such boundaries have been observed in simulations . but this in itself does not appear to imply that the dependence of the order parameter on growth rate must be continuous at the transition . the one - dimensional totally asymmetric exclusion process ( tasep ) , for example , exhibits a discontinuous phase transition in a certain region of parameter space , even though the position of the boundary between high- and low - density phases fluctuates over the entire system @xcite .
hinrichsen @xcite also reported simulation results supporting a continuous transition in the tcm at diffusion rate @xmath31 , that is , above the estimate for @xmath32 given in @xcite .
it is now generally acknowledged that @xmath33 in the tcm . in a recent study @xcite ,
park reported simulation results that support dp - like scaling in the tcm at diffusion rates 0.95 and 0.98 .
specifically , the order parameter ( starting from a filled lattice ) appears to decay at long times as @xmath34 , with @xmath35 taking its dp value , over about two decades in time .
we note however that the decay exponent is very sensitive to the choice of the time interval used for analysis and of the control parameter @xmath19 . analyzing simulation results for @xmath21 in studies extending to @xmath36 mcs
, we obtain local decay exponents @xmath37 between 0.1 and 0.2 , varying @xmath19 in a very narrow range . a crossover between a long supercritical plateau for @xmath38 and a rapid decay to an inactive state can not be ruled out . for @xmath21 , the regime during which dp - like scaling is found in ref .
@xcite ( i.e. , @xmath39 ) corresponds to overall particle densities in the range 0.67 - 0.32 . while scaling behavior can be observed at such densities in the contact process @xcite , definitive results for the decay exponent would require studying systems with substantially smaller values of @xmath18
. finally , three or more critical exponents would have to be determined to demonstrate convincingly that the transition falls in the dp class .
although the results of hinrichsen and of park do not appear to rule out rigorously a discontinuous transition in the tcm , we believe that they are fundamentally correct . this conclusion is based not on simulation results but rather on a mapping to a modified cdp process , to be developed in sec .
v. analysis of this mapping in leads to the conclusion that the transition is in fact continuous , despite abundant numerical evidence to the contrary .
one of the most common theoretical approaches to markov processes with spatial structure is a truncation of the master equation known as an @xmath2-site approximation @xcite .
such approximations have been applied to the tcm in efforts to determine the order of the transition ; in this section we review and extend these results .
the simplest method in this family is dynamic mean - field theory or the _ one - site approximation _ , in which the probability of an @xmath5-site configuration is factored into a product of @xmath5 single - site probabilities , so that , for example , @xmath40 where @xmath41 ( @xmath42 ) denotes an occupied ( vacant ) site and @xmath18 is the fraction of occupied sites . the resulting equation for @xmath43 yields rather poor predictions for the tcm ; better results
are obtained using larger clusters . in the @xmath2-site approximation ,
the equations that govern the probability distribution for clusters of @xmath2 sites are truncated by expressing the probabilities of @xmath44 site ( or larger ) clusters in terms of the @xmath2-site distribution . in the 3-site approximation , for example
, we write @xmath45 . as @xmath2 grows
, the number and complexity of the equations increases rapidly , but it is possible to generate the equations , and integrate them numerically , via a computational algorithm @xcite . in ref .
@xcite , the 4-site approximation for the tcm was found to predict a continuous phase transition for diffusion rates @xmath46 and discontinuous one for @xmath47 .
the predicted value for the tricritical diffusion rate @xmath32 , however , is much smaller than that reported in simulations ( @xmath48 ) .
since the phase diagram predicted by the @xmath2-site approximation generally converges to the correct one as @xmath49 , it is of interest to study the results for larger @xmath2 . in certain cases , predictions
based on a sequence of @xmath2-site approximations behave in a consistent manner , and can be extrapolated to provide estimates of the transition point and critical exponents , via the coherent anomaly method @xcite . for small @xmath2 ,
the position of @xmath32 varies considerably .
for example , the @xmath50 approximations yield a discontinuous transition even for @xmath51 , but for @xmath52 there is a tricritical point at some @xmath53 . the estimates for @xmath32 increase gradually with @xmath2
; for @xmath54 , one finds @xmath55 , for example . on the other hand , for a fixed , large diffusion rate ,
the transition remains discontinuous , with a large jump in the order parameter , which does not diminish appreciably with increasing @xmath2 . for @xmath21 , for example , fig .
[ cs_98 ] shows that all the approximations studied ( @xmath56 ) yield a discontinuous transition .
recently , ferreira and fontanari published results casting doubt on the utility of @xmath2-site approximations for the tcm @xcite .
they show , for example , that for @xmath51 , the values of @xmath25 veer _ away _ from the simulation value as @xmath2 is increased from 11 to 18 ; our studies confirm this observation .
( we note that a nonmonotonic approach to the critical point is observed in a stochastic cellular automaton in which at least three particles are required for particle generation or survival @xcite . )
moreover , the values of the tricritical reproduction rate @xmath57 ( for @xmath58 ) appear to converge to an unphysical ( negative ) value as @xmath49 @xcite . thus if the @xmath2-site approximations converge to the correct values , they do so in a nonmonotonic fashion , such that for the cluster sizes accessible with present technology ( @xmath59 or so ) , quantitative results for @xmath25 can not be obtained for the tcm . although our @xmath2-site approximations show stable nonvanishing gap sizes at @xmath21 for @xmath60 ( fig .
[ cs_98 ] ) , we find that for a given level ( @xmath61 , say ) , @xmath62 is not a monotonic function of @xmath17 , and the location of the tricritical point is rather uncertain .
[ cs13tr ] shows that on increasing the diffusion rate from @xmath63 to @xmath64 , the critical point shifts to higher values , but for @xmath65 this tendency reverses . on the other hand
, there is no evidence of a discontinuous transition for @xmath66 .
our analysis provides a higher @xmath67 estimate for @xmath61 than found in @xcite ; the reason for this difference is not known , since our result for @xmath68 agrees with that reported in the latter work .
we observe a similar behavior for @xmath69 and @xmath70 ; our tricritical point estimates do not fit on the extrapolation line given in @xcite . for
the cluster sizes studied , @xmath71 can not be fit with a linear function of @xmath72 , so that the @xmath73 limiting value can not be estimated with confidence .
it seems likely that @xmath71 exhibits an oscillatory convergence with @xmath2 ; if so , the question of whether @xmath74 , ( that is , the existence of a discontinuous transition ) , can not be resolved using the available @xmath2-site approximation results .
we study the tcm via monte carlo simulation , using several approaches that complement earlier analyses : dependence on the initial value of the order parameter , interface dynamics , scaling in the presence of a weak source of activity , and scaling of the quasistationary order parameter .
while some of the results would seem to provide good evidence of a discontinuous transition , we shall defer our conclusion until the following section .
if a phase transition is discontinuous , the evolution of the system should depend strongly on the initial condition , while at a continuous transition the evolution is toward the same qs state , regardless of the initial condition . in @xcite ,
simulations of the tcm at diffusion rate @xmath21 are reported , showing that at the transition ( @xmath75 ) , the value of the order parameter at long times depends on its initial value . for initial particle densities
@xmath76 between 0.3 and unity , the system evolves to the active state , while for @xmath77 it rapidly approaches the absorbing state .
( in studies using @xmath78 , the initially occupied sites are chosen at random , uniformly over the lattice . )
these results demonstrate that an active phase , characterized by a high value of the order parameter , is accessible starting from a high density , but not from a low one .
the findings for high diffusion rate are in sharp contrast to those found for @xmath51 , for which the critical reproduction rate is @xmath79 . in this case
, the particle density attains the same qs value , starting from very different initial values .
here we extend the simulations of @xcite to much larger systems .
we follow the evolution of the order parameter @xmath80 in systems of @xmath81 sites ( with periodic boundary conditions ) for times of up to @xmath36 mcs , averaging over 5 -20 realizations . for @xmath21
, we find that for low initial values ( @xmath82 ) , the density falls exponentially for @xmath83 ( see fig.[3410_98 ] ) . for the same reproduction rates , using @xmath84 , the order parameter @xmath80 exhibits a long plateau ( @xmath85 mcs ) at a high density . at longer times
@xmath80 decays , as expected in a finite system . for @xmath86 , by contrast , the order parameter curves @xmath80 , starting from high and low initial values , attain a common value at long times ( see fig.[3410_5 ] ) . in these studies
the initial configuration consists of two blocks , one fully occupied , of @xmath87 sites , and the other , of @xmath88 sites , completely empty . for @xmath21 and @xmath89 , one finds that in the initially occupied region , the particle density quickly relaxes to its qs value of about @xmath90 .
the ensuing evolution is characterized by the drift of the boundaries between active and empty regions .
a given realization stops either when it attains the absorbing state , or when the number of particles indicates that the active phase has filled the entire system ( we use a particle number of @xmath91 as the criterion for this event ) .
figure [ qcnc7 ] shows a typical history for @xmath92 and @xmath93 .
( the graph shows the mean density in blocks of 50 sites , with time increasing downward , in steps of 10000 time units between each density profile . )
the boundaries between active and inactive regions appear to follow independent , unbiased random walks .
if the boundaries can be represented by independent , unbiased random walkers , then the number @xmath94 of particles ( which is approximated by @xmath95 times the size of the active region ) , should also follow an unbiased random walk .
the walk starts at @xmath96 and is subject to absorbing frontiers at @xmath97 and @xmath98 .
well known results on random walks @xcite then imply that the probability @xmath99 of reaching @xmath100 before @xmath97 is given by @xmath101 .
we estimate @xmath99 in sets of 100 realizations , on rings of 500 , 1000 , and 2000 sites ; the linear trend evident in fig .
[ tcrnc ] supports the fluctuating boundary interpretation .
this in turn suggests that at high diffusion rates , the dynamics of active and inactive domains is effectively that of cdp .
away from the phase transition , we expect the interface to drift on the average , advancing into the inactive region for @xmath102 and vice - versa .
we determined the mean interface velocity for @xmath21 on rings of 3000 sites , using @xmath103 . after allowing the system to relax for 50@xmath104000 mcs , we record the density profile @xmath105 , and determine the interface position @xmath106 via the criterion @xmath107 , with @xmath108 the bulk particle density at the @xmath19 value of interest .
we then allow the system to evolve for an additional 20@xmath104000 mcs , and again determine the interface position .
[ wv3k ] shows the interface drift velocity @xmath109 , as determined in samples of @xmath110 realizations , varying linearly with @xmath26 ; linear regression yields @xmath111 for @xmath112 , in agreement with other estimates of the transition point .
( we verified that the interface velocity obtained using a ring of 5000 sites , and observation times of @xmath113 and @xmath114 mcs , agrees to within 2% with the value obtained using the smaller system . )
the above results show that for large @xmath17 the system is divided into well defined active and empty regions . a typical configuration of a large system ( @xmath115 sites ) at the transition ( @xmath21 , @xmath116 ) bears this out .
in order to visualize the configuration of a large system , we plot the cumulative particle number @xmath117 versus position @xmath118 , where @xmath119 is an indicator variable taking values of 0 and 1 at empty and occupied sites , respectively .
thus the local density @xmath120 corresponds to the slope of the graph at @xmath118 , with empty regions corresponding to horizontal lines . in these studies
we initially occupy half the sites , randomly , so that initially the local density is @xmath121 . in the initial phase of the evolution
, the global density rapidly grows to about 0.8 ; thereafter it begins to fluctuate , as empty regions form .
the configuration shown in fig .
[ sc2021 ] , for a time of about @xmath122 mcs ( comparable to the simulation times in @xcite ) , consists of a series active regions , with density @xmath123 , and empty regions , giving an overall density of 0.25 . of note
is the high density in the active regions , and the similarity of the density in active regions separated by large inactive gaps .
we verified that for this choice of @xmath17 and @xmath19 , the density in active regions @xmath124 , _ independent of system size and of overall density_. the system reaches the absorbing state via fluctuations of the boundaries between active and inactive regions , which eventually drive the active fraction to zero , while the active region density remains constant . for comparison , in fig .
[ sc2021 ] we also show a typical configuration for @xmath51 and @xmath125 .
the initial condition is the same as for @xmath21 , and the simulation is again halted when @xmath18 falls to 0.25 , which occurs at @xmath126 mcs for these parameters . in this case
the empty regions are typically much smaller than under rapid diffusion , and the local density in active regions varies considerably .
we turn next to studies of the stationary order parameter in the presence of a weak source , @xmath127 , defined as the probability , per unit time and per vacant site , to insert a particle . in @xcite
a weak source was used to demonstrate hysteresis ; here , we consider the scaling of the order parameter as @xmath128 at the transition . for @xmath51 ,
we determine @xmath18 at @xmath129 , varying @xmath127 between @xmath130 and @xmath131 .
for each @xmath127 value , a series of lattice sizes ( from 5000 to 50000 sites ) are used to estimate the limiting infinite - size value of @xmath18 .
we verify the scaling law @xmath132 , with @xmath133 , in agreement with the value expected for directed percolation in one spatial dimension , @xmath134 @xcite . for @xmath21 and @xmath135
, we observe a very different scenario . for a given value of @xmath127 ,
two values of @xmath18 are found , depending on the initial density .
for large initial densities , @xmath18 approaches a value of about 0.815 as @xmath128 , while for a low initial density , @xmath136 ( see fig .
[ tcrsd98 ] ) .
the order parameter , moreover , is essentially independent of system size for @xmath137 .
these results are consistent with a discontinuous transition , and the absence of critical scaling , for @xmath21 . to close this section
we report results on the qs value of the order parameter as a function of system size . in ref .
@xcite , estimates for the limiting ( @xmath138 ) value of @xmath95 were found to exhibit a discontinuity at the transition , for @xmath139 ; here we focus on @xmath21 .
we determined @xmath18 in qs simulations of duration @xmath140 mcs ( for @xmath141 ) up to @xmath142 mcs ( for @xmath143 ) , allowing the first 10% of the time for relaxation . in fig .
[ tcfs ] we plot the qs order parameter versus @xmath144 , for @xmath19 values near the transition .
the curves divide into two families .
( for @xmath145 ) approaches zero as @xmath138 , while for larger values of @xmath19 , the density approaches a nonzero limiting value .
in the minute interval @xmath146 the limiting ( @xmath138 ) value of the order parameter @xmath18 jumps from zero to about 0.6 .
the inset shows that at the transition point , @xmath147 , @xmath18 decays exponentially with system size . at a continuous transition one
expects the density to decay as a power - law , @xmath148 .
at various points in this discussion we have drawn an analogy between the tcm at high diffusion rates and compact directed percolation .
the picture that emerges from the simulations reported above , and from some of the earlier studies @xcite , is that in this regime the system contains essentially two kinds of regions , one of high density , the other empty .
the boundaries between these regions perform independent random walks , leading eventually to extinction of activity .
the above scenario corresponds to the phase transition in cdp , and suggests that we construct a reduced description in which blocks of @xmath149 sites in the active region ( with particle density on the order of @xmath150 ) correspond to sites in state 1 of the cdp , and blocks of @xmath149 empty sites correspond to sites in state 0 in the cdp .
the dynamics of the cdp consists exclusively of random walks performed by the interfaces between strings of 0s and strings of 1s . as we vary @xmath19 through its critical value in the tcm , the drift velocity of the interfaces in the corresponding cdp passes through zero , and the asymptotic density of 1s jumps from zero to one . if the above caricature of the tcm as an effective compact directed percolation model were valid , a discontinuous transition would be guaranteed .
there are , however , two additional processes that must be taken into account .
evidently , gaps ( strings of empty sites ) can arise within active regions , else the tcm starting from a fully occupied lattice would never reach the absorbing state .
a fundamental point is that the process of gap nucleation , while essential to the tcm dynamics , occurs at an extremely small rate . by
gap nucleation " we mean the generation of a gap large enough ( of @xmath151 sites , say ) that its boundaries fluctuate independently of one another .
gaps of size @xmath152 are equally likely to grow or to shrink , whereas smaller gaps tend to shrink , due to particles diffusing in from the adjacent occupied regions . the data of fig .
3 permit an order of magnitude estimate of the rate of gap nucleation : the density begins to fall appreciably from its plateau value at @xmath153 , and there are @xmath154 sites in the system , giving a rate of @xmath155 per site . in fig .
[ tgap ] we plot the mean first - passage time @xmath156 for the appearance of a gap of size @xmath157 in a system of @xmath158 sites ( parameters @xmath21 and @xmath116 ) , starting with all sites occupied .
( in these studies @xmath156 is estimated using samples of @xmath159 50 to 500 realizations , with smaller @xmath160 for larger system sizes . )
the first - passage time grows rapidly for smaller sizes and then crosses over to a slower growth around @xmath161 , at which point @xmath162 .
identifying this crossover size as @xmath151 gives a nucleation rate of @xmath163 per site .
it is natural to take the block size @xmath149 in the cdp mapping as the critical gap size @xmath151 , so that a one - site gap in the equivalent cdp process is equally likely to grow or to shrink to zero .
since we map @xmath151 tcm sites to a single site in the cdp , the effective gap nucleation rate in the latter is then @xmath164 .
the estimates for the nucleation rate and for @xmath151 given above yield @xmath165 in the range @xmath166 per cdp site . as noted above
, a small gap can shrink when particles diffuse in from outside .
can a large gap be destroyed in this manner ? to answer this , consider an interface ( treated as fixed , for the sake of this argument ) between large active and empty regions : the local density @xmath167 for @xmath168 and @xmath169 for @xmath170 .
think of the edge of an active region as a particle source .
particles are emitted at a rate of order @xmath171 , diffuse away at rate d , and decay at rate @xmath172 . in the stationary state ,
a continuum diffusion analysis yields a local density for @xmath170 of @xmath173 giving @xmath174 with an interface width @xmath175 . for @xmath21 and @xmath89
this gives @xmath176 ; for these parameters simulation shows that near the edge of a large gap , the density decays to zero @xmath177 .
nucleation of an active region inside a gap occurs at a rate @xmath178 ^ 3 $ ] , making the probability of nucleating activity deep within a gap negligible . for @xmath179 and @xmath17 and @xmath19 as above , for example , we find a nucleation rate of @xmath180 .
the essential point is that @xmath120 decays exponentially , so that nucleation of activity is limited to the neighborhood of the edges , whereas the nucleation of gaps can occur anywhere inside an active region . thus we conclude that the tcm at a high diffusion rate is equivalent to compact directed percolation with a very small , but nonzero rate of gap formation within clusters of 1s . but this process is in turn equivalent ( insofar as scaling properties are concerned ) to the domany - kinzel cellular automaton ( dkca ) @xcite with @xmath181 and @xmath182 .
( recall that in the one - dimensional dkca , @xmath183 is the probability of a site taking state 1 , given one neighbor in state 1 , and one in state 0 , at the preceding time , and that @xmath184 is the probability of state 1 given that both neighbors are in state 1 at the previous step . )
cdp corresponds to the line @xmath185 . from the work of janssen @xcite and of lbeck
@xcite we know that for @xmath186 , the transition occurs at @xmath187 .
the phase transition of the dkca is discontinuous _ only _ for @xmath188 ; for _ any _ @xmath189 , it is continuous and belongs to the dp universality class .
we are led to the same conclusion regarding the tcm : for any @xmath190 , there is a small but finite rate of nucleating gaps within active regions , so that the effective value of @xmath184 is slightly less than unity . given the nearness of the equivalent dkca to the line @xmath188 , it is not surprising that simulations of the tcm using lattice sizes @xmath191 and simulation times @xmath192 yield an apparently discontinuous phase transition , for @xmath193
. it is only for large systems and long simulation times that the effects of gap nucleation become apparent , as in the studies of @xcite and @xcite .
( indeed , the results shown in fig . 3 are also compatible with dp - like decay of the order parameter . ) but numerical studies of the stationary order parameter can be expected to show discontinuous behavior , this near the cdp line . for similar reasons ,
it is not surprising that @xmath2-site approximations , using clusters of fewer than twenty sites , miss the effect of gap nucleation ; much larger clusters would be needed to capture this properly .
it is perhaps worth recalling that cdp - like spreading behavior is observed in a _ surface modified _ version of directed percolation in one spatial dimension @xcite . in this case ,
propagation of activity at the edges of the active region occurs with a different creation rate ( @xmath194 , say ) , than in the bulk , which has a creation rate of @xmath195 .
let @xmath25 denote the critical creation rate for the original problem , that is , for @xmath196 .
for @xmath197 , the phase transition occurs at some bulk creation rate @xmath198 , which means that the bulk has a finite activity density even for @xmath199 .
thus the active cluster is compact , and the scaling behavior is that of cdp .
the essential difference between this model and the tcm is that in the former case large gaps can not be nucleated within the active region : since @xmath198 at the transition , gaps tend to shrink .
we have presented various pieces of evidence suggesting that the order parameter is a discontinuous function of the creation rate in the triplet creation model at high diffusion rates ( our numerical studies focus on @xmath21 ) .
we do not find evidence of stable coexisting active and inactive regions ; the boundaries between these regions are observed to fluctuate , as asserted in ref .
the numerical evidence in favor of a discontinuous transition includes hysteresis ( under a weak source ) , the initial - density dependence of the order parameter at later times ( with or without a particle source ) , the high particle density within active regions , the random - walk - like fluctuations of the boundaries between active and inactive regions , and the apparent absence of power - law scaling of the qs order parameter as a function of system size , at the transition point . despite the numerical evidence in favor of a discontinuous transition , a mapping of the tcm to a effective dynamics resembling that of compact directed percolation , but with a very small gap nucleation rate , leads to the conclusion that the transition is continuous for any @xmath200 .
the studies reported above , suggesting a discontinuous transition , were performed using relatively small systems and/or limited times .
for example , to see the effect of gap nucleation in the studies with a weak source of activity , the source strength @xmath127 would have to be much smaller than the gap nucleation rate @xmath201 .
the failure of @xmath2-site approximations to give a clear indication of the nature of the transition may again be attributed to the very small gap nucleation rate for @xmath202 .
these approximations , however , are problematic even for @xmath51 : the values predicted for @xmath25 do not converge monotonically to the correct value with increasing @xmath2 @xcite .
the continuous nature of the transition in the one - dimensional tcm was of course asserted some time ago by hinrichsen @xcite , and received further support in park s simulations @xcite .
our argument nevertheless contributes to an intuitive understanding of this result , and might provide the basis for a rigorous demonstration of the continuous nature of the transition .
we are grateful to jos f. fontanari and mrio j. de oliveira for helpful discussions .
support from hungarian research fund otka ( grant nos .
t046129 , t77629 ) is acknowledged .
the authors acknowledge access to the hungrid , clustergrid and the supercomputer of niif budapest .
rd acknowledges support from cnpq , and fapemig , brazil . by _ explicit _ diffusion we mean a particle hopping or exchange process , in contrast to _ implicit _ diffusion , generated by the dynamics of particle annihilation and creation at neighboring sites .
thus the coarse - grained description of the contact process , and many other models includes as diffusion term , despite the absence of explicit diffusion . | we study the lattice reaction diffusion model @xmath0 , @xmath1 ( triplet creation " ) using numerical simulations and @xmath2-site approximations .
the simulation results suggest that the phase transition is discontinuous at high diffusion rates . in this regime
the order parameter appears to be a discontinuous function of the creation rate ; no evidence of a stable interface between active and absorbing phases is found .
based on an effective mapping to a modified compact directed percolation process , we shall nevertheless argue that the transition is _ continuous _ , despite the seemingly discontinuous phase transition suggested by studies of finite systems . |
in low dimensional systems , the description of transport in terms of linear phenomenological laws , that works well in three dimensions , may not be possible .
one example is the fourier law which relates the thermal current density @xmath8 to the thermal gradient @xmath9 by @xmath10 , where the @xmath11 is the thermal conductivity of the material .
theoretical investigations of a number of translation invariant one - dimensional systems show that the fourier law is not followed in these systems @xcite .
the ratio of the current density to the thermal gradient is no longer independent of the system size .
nevertheless this ratio , again called ` thermal conductivity ' , is still useful in characterizing anomalous transport .
the calculation of ` thermal conductivity ' in these systems can be problematic .
ideally one would like to calculate the non - equilibrium stationary state ( nss ) measure of the system and calculate the expectation values of @xmath8 and @xmath9 using the nss measure .
this requires the existence of local thermal equilibrium ( lte ) so that a local temperature @xmath12 is properly defined .
a method generally used when considering a one dimensional system with its ends at two different temperatures @xmath13 and @xmath14 , is to calculate directly the current @xmath8 and the quantity @xmath15 , @xmath16 being the length of the system .
another approach is to use the kubo - green formula which involves the time integral of the equilibrium current - current correlation function
. the first formula does not require the existence of lte whereas the second does and it is not clear how the results from the two methods can be compared . in 1-d systems with translation invariance ,
the ` conductivity ' , @xmath11 , shows a power - law dependence on the size of the system ; @xmath17 @xcite .
a question that has been investigated extensively is whether @xmath18 is universal across a class of systems .
the other related questions are about the number of such universality classes and the values of @xmath18 characterizing them .
such an investigation has to take into account the difficulties posed by the considerations in the previous paragraph .
hydrodynamic mode - coupling theory predicts two universality classes corresponding to @xmath19 and @xmath20 for one - dimensional momentum conserving systems @xcite .
specific models of one - dimensional lattice systems with pair - wise nonlinear interactions between neighboring particles have been studied extensively using numerical and analytical methods @xcite .
the analytical studies involve approximations at some stage , whereas numerical studies are handicapped by the system size and the time scales that can be achieved in computers which makes it difficult to give a definitive answer to the question of the universality of @xmath18 .
results of these studies agree with the hydrodynamic results that the generic value of @xmath18 is @xmath21 ; when the leading order nonlinearity is cubic @xmath20 @xcite .
however , for the so - called fermi - pasta - ulam chain , where the nonlinearity is quartic , there is still no consensus ; numerical simulations show different values of @xmath18 @xcite and analytic calculations , both classical and quantum , show @xmath22 @xcite .
apart from the obvious theoretical interest in these problems there has also been a lot of experimental interest in studying heat transport at micro and nano scales @xcite .
recent experimental studies on nanowires such as carbon nanotubes and boron - nitride nanotubes show a clear breakdown of the fourier law @xcite . in order to attempt an understanding of realistic systems ,
it is essential to include all the modes of vibrations .
theoretical results , both numerical and analytical , mentioned above have been obtained for only the longitudinal vibrations of the chain .
it is natural to investigate how the exponent @xmath18 is affected by the transverse vibrations .
chains with both longitudinal and transverse vibrations have only been studied in the work of wang and li @xcite .
the wang - li model contains two kinds of intractions ; a pairwise interaction between the neighboring particles and another interaction that depends on the bending of the chain .
the harmonic approximation of the model yields modes whose frequencies @xmath23 depend on the wave - vector @xmath0 in the following manner . for small q
, the longitudinal modes have @xmath24 whereas the transverse modes @xmath25 .
wang and li used the mode - coupling analysis ( mca ) and molecular dynamics ( md ) simulations to study this system .
their first important observation from md simulations is that the phonon dispersions are strongly renormalized from their bare values , particularly the transverse phonon dispersion gets renormalized to become linear in q. for this reason they use an effective hamiltonian method in which the form of the hamiltonian and its parameters are taken from the md simulations . for the conductivity exponent
they obtain a generic value @xmath20 , but they also observe a value of @xmath22 in their md simulations which is attributed to crossover effects . when the parameter measuring the strength of transverse vibrations is sufficiently large , they also find that @xmath19 . in this paper , we study the quantum mechanical version of this model along the lines of @xcite . in this procedure ,
one essentially calculates the second order self - energy of the phonons self - consistently , which enables us to obtain the wave - vector dependent relaxation rate @xmath26 .
the thermal conductivity of the chain is then obtained through the use of kubo formula .
the finite size of the chain is taken into account by putting on the time integral in the kubo formula an upper cutoff proportional to the size of the chain @xcite .
this prescription yields @xmath27 , where @xmath28 governs the small - q behavior of the relaxation rate , @xmath29 . for the wang - li model , we find that in the second - order perturbation calculation , the relaxation rate @xmath30 for the transverse phonon diverges as @xmath31 , whereas the relaxation rate @xmath32 for the longitudinal phonons becomes constant .
this makes the bare phonons invalid normal modes .
these divergences basically arise due to the @xmath33-dispersion of the transverse modes .
this necessitates a self - consistent study of the phonon dispersion for this system @xcite .
such studies have been done on several systems , including one - dimensional systems as test cases for the procedure @xcite . here
we follow the method of gtze and michel @xcite which provides a nice connection between the mori - zwanzig mode coupling technique and the perturbation theory .
we show that this procedure indeed renormalizes self - consistently the transverse phonon dispersion to be linear , in qualitative agreement with the numerical results of md simulations @xcite .
we further note that the result on the renormalized frequencies is independent of the bare frequencies , but does depend on the fact that the cubic interaction vanishes linearly with q. since the vanishing of the interaction with any of its wave - vector arguments is a consequence of the translation invariance , this result should be generic to translation invariant one - dimensional systems .
we then perform a self - consistent analysis for the relaxation rates of the renormalized modes using the second order perturbation diagrams .
this procedure leads to integral equations from which small - q dependence of the relaxation rates @xmath34 can be extracted .
the main results of this analysis are : @xmath3 and @xmath4 .
gratifyingly , this behavior matches the results derived by wang and li for the classical system using the mode - coupling analysis , which arrives at the results through a seemingly different mechanism .
finally we use these results for understanding the behavior of conductivity with the system length @xmath16 .
this paper is organized in the following manner . in sec .
ii , we present the model and a calculation of the relaxation rates of the bare phonons . as mentioned above , this calculation shows that the modes are not well defined .
accordingly in sec .
iii , we present a self - consistent calculation for the frequencies of the modes , which shows that both the modes should disperse linearly at small wave - vectors . in sec .
iv , we calculate the relaxation rates of the renormalized phonons using second order perturbation theory . in section
v , we formulate equations for the relaxation rates using the self - consistent second order perturbation theory and obtain their behavior at small wave - vectors . in sec .
vi , we turn to the calculation of thermal conductivity using the finite size procedure mentioned above .
we close the paper with a summary and discussion of our results in sec .
this model consists of a chain of equal mass point particles with two kinds of interactions .
first is a pairwise interaction between nearest - neighbor particles which depends only on the absolute value of the distance between the particles and the second is a three particle interaction between neighbors that takes into account the bending of the chain .
there are two transverse directions , but due to the axial symmetry of the chain the two directions are entirely equivalent .
accordingly for understanding the qualitative role of the transverse vibrations on the relaxation of modes and transport , it suffices to consider just one transverse mode . in terms of the position @xmath35 and momentum @xmath36 vectors ,
the hamiltonian is written as @xmath37,\ ] ] where @xmath38 is the lattice constant , @xmath39 and @xmath40 is the unit vector along @xmath41 .
the mass of the particles has been set to unity .
this is a complicated many body problem .
one assumes that a few lower - order terms in the taylor expansion of the nonlinear potential is enough to give the generic physical picture .
also the low - temperature transport properties of the system are expected to be described by the low - energy modes of the system .
so we set up the problem in terms of phonon modes . denoting the deviation from equilibrium position of the particles by @xmath42 we have @xmath43 and @xmath44 .
fourier transform along the length of the chain is defined by @xmath45 where @xmath6 is the number of particles in the chain .
the hamiltonian , considering nonlinearities up to cubic terms only , can be written as @xcite @xmath46 where the summation indices in bold letters indicate that the summation is over all the polarization indices @xmath47 which takes values from @xmath48 and over all the wave - vectors @xmath49 , a notation we will follow hereafter . @xmath50 s are symmetric under exchange of pairs @xmath51 .
translation invariance imposes two conditions on @xmath50 ; they vanish when any of the wave - vectors is set to zero and @xmath52 , where @xmath53 is unity when @xmath54 is a reciprocal lattice vector and zero otherwise .
explicit forms for the @xmath50 s are obtained by expanding the potential term in eq .
( [ hamiltonianfull ] ) in taylor series and taking the fourier transforms .
the harmonic terms are given by @xcite @xmath55 note that the frequency of the longitudinal mode vanishes linearly with the wave - vector whereas for the transverse mode it vanishes quadratically .
henceforth we choose units in which @xmath56 .
the reflection symmetry in the transverse direction permits only two kinds of cubic terms to be nonzero , these being @xmath57 and @xmath58 .
both these interactions are proportional to @xmath59 .
these forms are generic when the potentials are taken to be functions of absolute values of relative displacements ( @xmath60)@xcite . for the fermi - pasta - ulam chain @xcite in which only the longitudinal vibrations are considered , the perturbation analysis is done around the bare phonon modes as they are well defined . but in this model we shall see that the bare phonon modes are not well defined , as the relaxation rates of the two kinds of phonons are not zero even for zero wave - vector mode .
this requires that one works with the renormalized phonons as discussed below .
we begin by calculating the relaxation rates of the bare phonons . for this purpose
we first define the bare phonon operators @xmath61 in terms of which the quadratic part of the hamiltonian becomes diagonal .
the hamiltonian can be written as @xmath62 where we have defined @xmath63 and @xmath64 . .
the bare phonon green s functions @xmath65 for the longitudinal phonons and @xmath66 for the transverse phonons are represented respectively by the solid lines and the dashed lines .
the diagrams ( a ) and ( b ) are two contributions for the longitudinal phonons , while the diagram ( c ) gives the sole contribution for the transverse phonons.,width=566 ] for a system with two phonon polarizations , in general one defines the single particle green s functions as the time - ordered expectation values @xmath67 \right\rangle$ ] , where the time dependence of operators is given by @xmath68 , @xmath69 being the imaginary time .
the angular brackets correspond to thermal averages defined by @xmath70/tr[e^{-\beta h}]$ ] .
the @xmath71 s are thus 2x2 matrices in polarization indices , but in the present case , the reflection symmetry in the transverse direction allows only the diagonal components to be nonzero . in the following
we just write them as @xmath72 .
the perturbation series for @xmath73 can be summed into the form @xmath74 , where @xmath75 s denote the usual matsubara frequencies and @xmath76 is the self - energy .
the relaxation rate is defined as @xmath77 , and the on - shell relaxation rate is denoted by @xmath78 .
we first evaluate the relaxation rate up to second - order in the perturbation series .
the second - order diagrams for @xmath79 that contribute to the relaxation rate are given in figs .
[ selfenergydiagrams ] ( a ) and ( b ) .
their contribution is @xmath80 where @xmath81 is the following unperturbed green s function . @xmath82 where @xmath83 is the bose factor and @xmath84 .
the relaxation rate is given by @xmath85 the sum is simplified by two conditions : the @xmath86-function contained in the potential @xmath87 and the @xmath28-function .
therefore , we can write @xmath88 where @xmath89 is the solution of @xmath90 and the jacobian @xmath91 . if the sum on the rhs of eq .
( [ gamm2barephonon ] ) is finite we expect @xmath92 , as @xmath93 .
we first note that the term corresponding to @xmath94 ( fig .
1(a ) ) which involves only the longitudinal modes gives zero contribution as the condition @xmath95 can have only trivial solutions at which the self - energy itself becomes zero @xcite . the second term corresponding to @xmath96 ( fig .
1(b ) ) needs to be examined for singular behavior .
the source for singularities are the jacobian @xmath97 and the number factors which for low frequencies have the form @xmath98 . for this term
we have transverse mode frequencies which vanish quadratically with the wave - vector and hence are more singular .
considering @xmath99 term , the small-@xmath0 solution is obtained to be @xmath100 and the jacobian @xmath101 .
this yields a contribution to @xmath32 which diverges for small @xmath0 at finite temperatures and we have the leading behavior @xmath102 .
the second - order diagram for @xmath103 is given in fig .
the yields the relaxation rate as @xmath104 proceeding as in the analysis of eq .
( [ gamm2barephonon ] ) , we see that the term corresponding to @xmath105 is finite for @xmath106 .
the above analysis shows that the bare phonon modes are not stable under scattering and thus are not good zeroth order approximations in the perturbation theory .
therefore we follow the mode - coupling method as formulated by gtze and michel @xcite to study how the bare phonon dispersions get affected by the nonlinearities . for completeness
we first present a brief outline of the mode - coupling method @xcite .
the procedure requires evaluation of scalar products defined in the space of the dynamical operators ( denoted by @xmath107 ) as follows .
@xmath108 \right \rangle , \quad\quad \im z\gtrless0 , \end{aligned}\ ] ] where @xmath109 . for us the key quantity is the static susceptibility @xmath110 .
it is easily seen that in the harmonic approximation @xmath111 .
the renormalized frequencies are the eigenvalues of the matrix @xmath112 for the nonlinear problem .
it follows from the definition and the space reflection symmetry of the hamiltonian that @xmath112 is a real symmetric matrix which is an even function of @xmath0 .
therefore , @xmath113 is diagonalizable with a complete and orthonormal set of eigenvectors @xmath114 obeying @xmath115 where the eigenvalues obey @xmath116 due to time - reversal symmetry . in order to develop equations for these scalar products ,
we utilize the identity , @xmath117 \rangle$ ] which follows from eq .
( [ chidefinition ] ) .
the lhs of this equation can be calculated using @xmath118 $ ] .
for further calculations we define @xmath119 . taking ( i ) @xmath120 and ( ii ) @xmath121 respectively , we get the following equations which are exact .
@xmath122 we write these equations in the matrix form , suppressing the @xmath0-dependence , as @xmath123^{-1 } \v{v},\ ] ] where @xmath124 and @xmath125 . to get a closure for this set of equations , we need to make some approximations .
we do this in a self - consistent manner by expressing the higher order correlation functions @xmath126 in terms of @xmath113 matrices .
the expression for @xmath127 is is given by @xmath128 the expectation contained in the above equation is evaluated using an effective quadratic hamiltonian given by @xmath129,\end{aligned}\ ] ] where the operators @xmath130 are the bosonic operators defined in the usual manner as @xmath131.\ ] ] here the frequencies @xmath132 and the polarization vector @xmath133 are defined in eq .
( [ renorm ] ) , and are to be determined self - consistently . under this approximation using eq .
( [ chidefinition ] ) we get @xmath134 where we have used the compact notation @xmath135 . further @xmath136 and @xmath137 .
( [ dequation ] ) and ( [ aaeqexpanded ] ) together provide us the self - consistent equations for determining the renormalized frequencies @xmath138 .
in order to obtain their solution , we first iterate them . starting with a diagonal @xmath139 in the rhs of eq .
( [ aaeqexpanded ] ) , we have @xmath140 and we see that only the diagonal terms survive . so this procedure gives a diagonal @xmath113 back if our starting @xmath113 is diagonal .
putting this information in eq .
( [ dequation ] ) reduces it to @xmath141 in order to analyze the low wave - vector @xmath142 behavior of @xmath143 , we simplify the equations by noting that @xmath144 we first consider @xmath145 , whose expression contains two terms corresponding to scattering by longitudinal ( @xmath94 ) and transverse ( @xmath96 ) modes respectively .
we may first suppose that the renormalized frequencies have the same @xmath0-dependence as the bare frequencies , @xmath146 .
then the use of eq .
( [ approxvn ] ) in eq .
( [ deqdiagonal ] ) shows that , as @xmath106 , the longitudinal contribution to @xmath145 behaves as @xmath33 whereas the transverse part gives rise to a divergence basically due to @xmath33-dependence of the transverse frequencies in the denominator .
similarly the expression for @xmath147 also involves transverse mode frequencies in the denominator , which again give rise to a singular dependence on @xmath0 .
if on the other hand we assume @xmath148 , all the integrals converge and we get @xmath149 . thus we have a self - consistent solution for @xmath150 in the @xmath106 limit , which is also diagonal .
this is exactly what is seen in the md simulations of the classical system with the wang - li interaction @xcite . from this analysis
it is clear that the linear solution obtained here is independent of the form of bare frequencies .
thus the linear @xmath0-dependence of the frequencies in one dimension seems to be a generic result . for further analysis
we take a specific form for renormalized phonon frequencies , namely , @xmath151 which agrees with the self - consistent solution in the @xmath106 limit .
we now study an effective hamiltonian given by eq .
( [ hcubicinnumberop ] ) and eq .
( [ adefinition ] ) in which @xmath152 is replaced by @xmath153 . the second - order relaxation rate for the renormalized modes are as before , @xmath154 first we estimate the low-@xmath0 form of @xmath155 .
this proceeds exactly as the analysis of eq .
( [ gamm2barephonon ] ) .
we have @xmath156 where @xmath157 are the solutions of @xmath158 and the jacobian is given by @xmath159 . since both @xmath160 the factors of @xmath0 in the denominator that arise due to number factors are canceled by the @xmath161 term . therefore if the jacobian @xmath162 does not go to zero at some @xmath157 , we get @xmath163 . there are four cases for the energy - momentum conservation conditions due to four possible values of @xmath164 and @xmath165 .
these have to be analyzed separately .
we choose the brillouin zone @xmath166 $ ] . for @xmath99 ,
the energy conservation condition becomes @xmath167 .
when @xmath168 there are two solutions @xmath169 and @xmath170 . but these correspond to no scattering and have to be excluded from the perturbation terms .
thus as noted earlier the relaxation rate within this approximation is zero @xcite . when @xmath171 nontrivial solutions are possible .
[ figomegaequation](a ) shows plots of @xmath153 and @xmath172 with respect to @xmath0 for a representative set of values for @xmath54 and @xmath173 .
this figure shows that , when @xmath174 , @xmath54 is a solution for the energy conservation condition @xmath175 when the value of @xmath0 corresponds to one of the intersections of the two graphs . for small-@xmath0
we get a solution @xmath176 with a finite value for the jacobian of the argument of the @xmath28-function .
therefore we see that the contribution to @xmath177 from this case is proportional to @xmath33 if @xmath174 and zero otherwise . ) can be satisfied .
the figure on the left shows @xmath153 and @xmath172 in the same graph for @xmath178 and the figure on the right shows @xmath153 and @xmath179 for @xmath180 .
dotted vertical lines mark the values of @xmath54 which is the solution for @xmath0 values corresponding to the intersections of the two graphs.,width=566 ] for @xmath181 , the @xmath28-function condition is @xmath182 .
[ figomegaequation](b ) shows lhs and rhs of this condition on the same graph .
we see that for @xmath183 we have a nontrivial solution and in the small-@xmath0 limit we find a solution @xmath176 where the jacobian also is finite .
therefore , for this case , the contribution to @xmath177 is proportional to @xmath33 for @xmath184 and zero otherwise .
next we take @xmath185 .
the energy conservation is given by @xmath186 .
we have a nontrivial solution for @xmath171 which for @xmath187 is @xmath176 .
again we find the contribution to @xmath177 is like @xmath33 for @xmath174 and zero otherwise .
the fourth case @xmath188 has only trivial solutions .
therefore , we have @xmath189 now we consider the relaxation rate @xmath190 .
it has contributions coming from both transverse and longitudinal modes corresponding to @xmath96 and @xmath94 in eq .
( [ gammxrenormalized ] ) .
the @xmath96 term again involves interaction between two @xmath191-phonons and one @xmath192-phonon as in the case for @xmath193 and its low wave - vetor behavior can be calculated in a similar way . as in the calculation of @xmath193 , for @xmath194 , there are nontrivial solutions satisfying the @xmath28-function except for @xmath195 . thus the corresponding contribution to @xmath196 is proportional to @xmath33 .
the @xmath94 term involves three @xmath192-phonons and the contribution to the relaxation rate is zero because the @xmath28-function condition @xmath197 has no nontrivial solution .
therefore @xmath198 thus , the renormalized phonon modes at low wave - vectors are slow decaying and well - defined to perform the perturbation analysis .
the phonon relaxation rates of a chain with cubic nonlinearity but with only the longitudinal vibrations was studied earlier by us @xcite . for this case , as mentioned in the above paragraph , the relaxation rate is zero in the second order as the two conservation conditions for energy and wave - vector can not be satisfied simultaneously . to overcome this difficulty we performed a self - consistent approximation which consists in replacing bare phonon green s functions in the second - order diagram ( fig.[selfenergydiagrams](a ) ) by full green s functions .
this relaxes the strict energy conservation condition and allows for nonzero solutions for the relaxation rate through a solution of an integral equation .
our aim in this section is to perform a similar analysis for the present case . for this purpose ,
one uses the spectral representation of the green s function , @xmath199 with the spectral function is given by @xmath200 where @xmath201 and @xmath202 . if we now replace @xmath203 by @xmath204 in eq .
( [ sigma ] ) , the corresponding equations for relaxation rate , eqs .
( [ gammxrenormalized ] ) and ( [ gammyrenormalized ] ) , are replaced by a set of self - consistent equations for @xmath205 and @xmath206 .
we get @xmath207 we follow ref @xcite in the analysis of these equations .
we note that for small @xmath0 , the relaxation rate @xmath1 is very small and the spectral function @xmath208 can be approximated by a narrow peak at @xmath209 .
also , @xmath210 .
further we may replace @xmath211 where @xmath212 is a smooth function of @xmath213 .
we use this approximation for the number factors @xmath214 in eq .
( [ selfconstfirst ] ) .
we also replace @xmath215 and @xmath216 in the expression for @xmath208 to be used in eq .
( [ selfconstfirst ] ) .
after these approximations , only the spectral functions depend on @xmath213 in eq .
( [ selfconstfirst ] ) and the integral can be performed using @xmath217}{((a - b)^2+\gamma_1 ^ 2+\gamma_2 ^ 2)^2 - 4\gamma_1 ^ 2\gamma_2 ^ 2}.\end{aligned}\ ] ] we note that @xmath218 because it depends on @xmath0 through @xmath153 which is even in @xmath0 and through @xmath87 factors which enters the perturbation terms as even function of @xmath0 only .
then , @xmath219 and @xmath220 .
we can write @xmath221 , with @xmath222 \nonumber \\ & & \times s_{+-}(qx , k i , q+k\,i),\end{aligned}\ ] ] where @xmath223}{[\chi^2_{s_1s_2}(k_1i_1,k_2i_2,k_3i_3 ) + \gamma^2_{i_2}(k_2)+\gamma^2_{i_3}(k_3)]^2 - 4\gamma^2_{i_2}(k_2)\gamma^2_{i_3}(k_3)}.\end{aligned}\ ] ] here we have introduced the definition : @xmath224 .
similarly , for the transverse modes we get @xmath225|v_3(-qy ,- kx , q+k\,y)|^2 s_{+-}(qy , kx , q+k\,y ) \nonumber \\ & & + [ 1+n(\omr_x(k))]n(\omr_y(q - k ) ) |v_3(-qy , kx , q - k\,y)|^2 s_{-+}(qy , kx , q - k\,y ) \big\}.\end{aligned}\ ] ] equations ( [ selfconstgxa ] ) , ( [ selfconstgxb ] ) and ( [ selfconstgy ] ) have to solved self - consistently to obtain the small wave - vector behavior of the relaxation rates .
we show that these equations have a solution @xmath226 the key feature that enables the solution of these equations is that for small @xmath0 , each of the equations can be written in the following schematic form @xmath227 here one factor of @xmath228 comes from the prefactor , which is basically a detailed balance factor and the other from the factor of @xmath229 .
the integrand @xmath230 is not known as it as it depends on @xmath231 s .
our strategy is to examine these integrands in eqs .
( [ selfconstgxa ] ) , ( [ selfconstgxb ] ) , ( [ selfconstgy ] ) at @xmath232 , by first assuming that @xmath233 . if the integral turns out to be finite , clearly @xmath234 is a self - consistent solution .
if on the other hand the integral diverges , it presumably behaves for small @xmath0 as @xmath235 .
so we assume that @xmath236 with @xmath237 , and examine the integral for small @xmath0 .
a solution is achieved if a self - consistent value for @xmath28 can be obtained . in examining the singularities of the integrand in eq .
( [ fkq ] ) , a further simplification comes from noting that the bose factors @xmath238 give rise to factors @xmath239 for small frequencies , but these are canceled by the similar factors in interaction term @xmath240 .
so all the singular behavior of the integrands come from the @xmath241-factors introduced in eq .
( [ sfactor ] ) .
first , let us consider eq .
( [ selfconstgy ] ) , the equation for @xmath30 . at @xmath232 ,
the @xmath241-factors could give rise to singularities at @xmath169 if we set @xmath242 .
for @xmath243 , in all the four terms in eq .
[ selfconstgy ] , @xmath244 for small @xmath54 ; for example , in the third term , @xmath245 . using eq .
( [ gammaselfconst ] ) one finds that @xmath246 for small @xmath54 . since the integral is finite at @xmath232 , @xmath247 .
we next consider @xmath32 which has been split into two terms @xmath248 and @xmath249 in eqs .
( [ selfconstgxa ] ) and ( [ selfconstgxb ] ) .
the analysis of @xmath248 proceeds in a similar way as @xmath250 as @xmath251 in both the terms of eq .
( [ selfconstgxa ] ) and we get @xmath252
. something different happens in eq .
( [ selfconstgxb ] ) for @xmath249 .
it has again two terms for @xmath253 .
the term @xmath94 involves only the longitudinal mode , and this situation is identical to the fpu-@xmath18 problem discussed in ref .
we note that @xmath254\propto q k^2 $ ] as @xmath255 . using eq .
( [ gammaselfconst ] ) , for small @xmath0 and @xmath54 , we get @xmath256^{-1}$ ] . at @xmath232 ,
the corresponding integral is divergent .
but the small @xmath0 divergence of the integral can be obtained by noting that the major contribution to the integral comes from small @xmath54 and small @xmath257 region . denoting the corresponding contribution to @xmath258 by @xmath259
, we get , following @xcite , @xmath260 as is easily checked , eq .
( [ gamma1 ] ) is solved by taking @xmath261 . for the @xmath96 term
@xmath262 $ ] for small @xmath0 and @xmath54 .
the leading order behavior of @xmath241 in this limit is obtained using @xmath263 .
we see that @xmath264 is not singular at @xmath169 for finite @xmath0 , @xmath265 .
thus the limit @xmath266 of this sequence is finite and the corresponding contribution to @xmath267 is proportional to @xmath33 .
we can now go back to the earlier equations and check if by setting @xmath268 .
the earlier results for @xmath30 and @xmath248 are affected .
the examination of the integrals shows that they remain finite .
therefore to the leading order , @xmath269 and @xmath247 establishing the solution given by eq .
( [ gammaselfconst ] ) .
in this section , we calculate the thermal conductivity of the system using the kubo - green formula . defining the thermal green s function for the total current operator @xmath97 as
, @xmath270 \right \rangle,\ ] ] the kubo - green formula for thermal conductivity can be written as @xcite @xmath271 where @xmath75 are the matsubara frequencies .
the current operator , up to leading order , is given by @xmath272 where @xmath273 , @xmath274 .
further , we neglect the cross - correlation between the x- and y- components of the current . under the approximation used in @xcite that vertex corrections to the two particle green s function eq .
( [ defw ] ) has the same wave - vector dependence as @xmath275 for small @xmath54 , one arrives at the kinetic theory formula for conductivity , @xmath276 where @xmath277 is the relaxation time of single particle green s function and @xmath278 is proportional to the mode specific heat .
we note that @xmath279 is well behaved for small @xmath0 and any possible divergence can come from @xmath280 only . the formula given by eq .
( [ conductivitykinetictheoryformula ] ) is for an infinite chain and this quantity is divergent because @xmath281 has a non integrable singularity at @xmath232 . to obtain the conductivity @xmath5 for large but finite size chain of @xmath6 particles
, we follow the prescription used by other workers in the area @xcite . in this procedure one
rewrites eq .
( [ conductivitykinetictheoryformula ] ) as a time integral in the following way .
@xmath282 since @xmath283 , we get @xmath284 .
one argues that when the boundaries of the system are connected to thermal baths , the correlation @xmath285 dies after a time scale of order @xmath6 due to interaction with the baths , i.e. the typical time it takes for the modes to travel across the chain . using in the time integral in eq .
( [ kuboformulanoneq ] ) such a cutoff yields @xmath286 it is to be noticed that the leading order divergence arise from the term corresponding to the @xmath191-modes . if the conduction was normal , conductivity would be finite and the contribution from these modes would be negligible compared to the contribution from the @xmath192-modes when the strength of the interaction giving rise to transverse oscillations , @xmath287 , is small .
but here the conductivity itself is a divergent quantity and its divergence is determined by the slower relaxation of the transverse modes compared to the longitudinal modes .
we have studied thermal transport in a chain of coupled particles in which particles can have longitudinal as well as transverse vibrations .
particles interact via anharmonic potentials proposed by wang and li @xcite . for this model ,
the mode frequencies in the harmonic approximation have the following dispersions at small wave - vectors : for the longitudinal mode @xmath288 ; for the transverse mode @xmath25 , where @xmath0 denotes the wave - vector .
we first show that due to quadratic dispersion of the transverse mode , the relaxation rates for both the modes become too large to leave them as meaningful quasi - particles .
so we calculate their renormalized frequencies taking into account the anharmonic interactions in a self - consistent approximation .
we find that this procedure yields the linear dispersion for both kinds of modes irrespective of their bare forms .
we then perform a calculation of the relaxation rates for the renormalized phonon in a self - consistent approximation based on the second - order perturbation diagrams .
this calculation yields that for the longitudinal phonons , the relaxation rate @xmath3 , while for the transverse mode @xmath289 .
we then use these results to calculate the thermal conductivity @xmath5 of a chain of @xmath6 particles .
the finite size is incorporated by using a cutoff of the order of @xmath6 in the time integral over current - current correlation function that occurs in the kubo formula for the conductivity .
this yields a contribution to @xmath5 from longitudinal phonons to be @xmath290 , that from transverse phonons to be @xmath291 .
the latter contribution dominates giving @xmath7 for large n. 99 f. bonetto , j. l. lebowitz and l. rey - ballet in _ mathematical physics _
( world scientific , london 2000 ) , p. 128 .
s. lepri , r. livi and a. politi , phys .
* 377 * , 1 ( 2003 ) .
g. r. lee - dadswell , b. g. nickel and c. g. gray , phys .
e * 72 * , 031202 ( 2005 ) . o. narayan and s. ramaswamy , phys .
lett . * 69 * , 200601 ( 2002 ) . santhosh .
g and deepak kumar , phys .
e * 77 * , 011113 ( 2008 ) .
s. lepri , r. livi and a. politi , europhys . lett . * 43 * , 271 ( 1998 ) .
t. mai , a. dhar and o. narayan , phys .
. lett . * 98 * , 184301 ( 2007 ) .
a. perverzev , phys .
e * 68 * , 056124 ( 2003 ) .
s. lepri , phys .
e * 58 * , 7165 ( 1998 ) .
g and deepak kumar , phys .
e * 76 * , 021105 ( 2007 ) .
s. volz , _ microscale and nanoscale heat transfer _ , topics in applied physics ( springer , new york , 2007 ) , vol . 107 . c. w. chang , d. okawa , h. garcia , a. majumdar and a. zettle , phys .
* 101 * , 075903 ( 2008 ) .
p. kim , l. shi , a. majumdar and p. l. mceuen , phys .
87 * , 215502 ( 2001 ) .
j. hone , m. whitney , c. piskoti , and a. zettle , phys .
b * 59 * , r2514 ( 1999 ) . c. w. chang , a. m. fennimore , a. afanasiev , d. okawa , t.ikuno , h. garcia , a. majumdar and a. zettle , phys . rev .
lett . * 97 * , 085901 ( 2006 ) .
j. s. wang and b. li , phys .
e * 70 * , 021204 ( 2004 ) .
a. a. maradudin and a. e. fein , phys . rev . *
128 * , 2589 ( 1962 ) .
w. gtze and k. h. michel , _ dynamical properties of solids _ edited by g. k. horton and a. a. maradudin , north - holland publishing company , 1974 , vol .
1 , chapter 9 .
h. horner , _ dynamical properties of solids _ edited by g. k. horton and a. a. maradudin , north - holland publishing company , 1974 , vol .
1 , chapter 8 .
g. k. horton and e. r. cowley , _ physics of phonons _
, ( springer , karpacz 1987 ) , p50 .
a. cuccoli , v. tongetti , a. a. maradudin , a. r. mcgurn and r. vaia , phys .
b * 46 * , 8839 ( 1992 ) .
e. s. freidkin , g. k. horton and e. r. cowley , phys .
b * 52 * , 3322 ( 1995 ) .
j. scheipers and w. schirmacher , z. phys .
b * 103 * , 547 ( 1997 ) . g. d. mahan , _ many - particle physics , third edition _ , ( kluwer academic / plenum publishers , new york ) p | we study thermal transport in a chain of coupled atoms , which can vibrate in longitudinal as well as transverse directions .
the particles interact through anharmonic potentials upto cubic order .
the problem is treated quantum mechanically .
we first calculate the phonon frequencies self - consistently taking into account the anharmonic interactions .
we show that for all the modes , frequencies must have linear dispersion with wave - vector @xmath0 for small @xmath0 irrespective of their bare dispersions .
we then calculate the phonon relaxation rates @xmath1 , where @xmath2 is the polarization index of the mode , in a self - consistent approximation based on second order perturbation diagrams .
we find that the relaxation rate for the longitudinal phonon , @xmath3 , while that for the transverse phonon @xmath4 .
the consequence of these results on the thermal conductivity @xmath5 of a chain of @xmath6 particles is that @xmath7 .
v#1 # 1#1^ |
laser cooled trapped ions offer a very high level of control , both of their motional and internal quantum states . at the same time , the large charge - to - mass ratio of ions makes their motion very sensitive to electric fields , both static and oscillatory .
thus , trapped ions recently emerged as a tool in small - force sensing@xcite .
more common applications of trapped ions are in quantum information science @xcite and frequency metrology @xcite .
all these applications can benefit from scalable ion - trap architectures based on microfabricated ion traps . in particular , a promising route to achieve scalable quantum information processing uses complex electrode structures@xcite .
considerable effort is made in developing microfabricated trap architectures on which all trap electrodes lie within one plane @xcite .
these so - called planar traps facilitate creation of complex electrode structures and are , in principle , scalable to large numbers of electrodes .
moreover , this approach makes use of mature microfabrication technologies and is ideally suited to approaches involving hybrid ion - trap or solid state systems @xcite . despite the advantages of planar trap architectures , a number of issues remain unsolved . to achieve reasonably large trap frequencies ,
planar traps require shorter ion - electrode distances than conventional three - dimensional traps @xcite .
this results in high motional heating rates for the ions @xcite and causes charge buildup via stray light hitting the trap electrodes@xcite .
in addition , the proximity of the charges increases the effect of charge buildup as compared to macroscopic three dimensional traps .
finally , planar traps do not shield stray electrostatic fields from the environment surrounding the trap as well as the three dimensional trap geometries tend to do .
combined , these effects make the operation of planar traps much more sensitive to uncontrolled charging effects . to harness the full advantages of segmented ion traps , ion - string splitting and ion shuttling operations are required@xcite . for the reliable performance of these operations , control of the electrostatic environment over the full trapping region
is necessary .
typically one employs numerical electrostatic solvers to determine the potential experienced by the ions and generates electrode voltage sequences that will perform the desired ion shuttling @xcite .
stray electrostatic fields , however , displace the ions from the rf - null of the trap and thus introduce so - called micromotion@xcite sometimes to the point where trapping is no longer feasible .
thus , precise characterization and compensation of stray electric fields in the trapping region is required .
conventional methods to sense and compensate the electric stray fields can not easily be extended to planar traps because typically the stray fields are quantified via the the doppler shift induced by the micromotion .
it is undesirable to scatter uv light from the trap electrodes , and , thus , for planar traps , the detection laser typically does not have a sizable projection on the motion perpendicular to the plane of the trap .
we address these issues by applying a new method to compensate for stray fields well suited for planar trap geometries @xcite . based on the voltages required to compensate the stray fields , we realize a single - ion electric field sensor characterizing the electric stray fields along the trap axis . we observe a strong buildup of stray charges around the loading region on the trap .
we also find that the profile of the stray field remains constant over a time span of a few months .
the strength of the electric stray fields and its position on the trap is correlated with the high heating rates observed close to the loading region @xcite .
we use a planar trap with gold electrodes deposited on a sapphire substrate to trap single @xmath0ca@xmath1ions at a height of 240 @xmath2 m above the trap plane , see fig.[fig : trap ] .
ions are created via two step photoionization from a neutral calcium beam using 250 mw/@xmath3 of laser light at 422 nm and 750 mw/@xmath3 of laser light at 375 nm .
both the laser beams are focused to a waist size of 50 @xmath2 m .
great care has been taken to minimize exposure of the trap surface to the neutral calcium beam .
schematic of the trap used for the measurements@xcite .
the dc electrodes are drawn in blue , the rf electrode in orange , and the ground plane in gray .
details of the bonding pads to the dc electrodes are not shown for simplicity .
the axes indicate the origin of the coordinate system .
the green line along the z axis on the central dc electrode indicates the range of axial positions in which the stray electric fields shown in fig.[fig : el - field ] were measured .
the circular mark on this line indicates the location used as a loading region , around which the highest increase in stray electric fields was observed.,scaledwidth=40.0% ] the rf electrode is driven at a frequency @xmath4 15 mhz , amplified to @xmath5mw and stepped up via a helical resonator in a quarter wave configuration to a voltage of approximately 100 v amplitude . a 2:1 asymmetry in the width of the rf electrode results in a tilt of the radio frequency quadrupole by approximately @xmath6 in the @xmath7 plane .
the dc electrodes are used to move the ion along the axial direction and to compensate the stray fields .
the dc voltages used for trapping and compensation are between -10 v and 15 v. typical secular frequencies in this work were @xmath8 ( 1.2 , 1.4 , 0.4 ) mhz where the primes refer to the frame of reference rotated by @xmath6 . for doppler cooling and detecting the ions
, we use a diode laser at 794 nm , which is frequency doubled using a ring cavity to produce a wavelength of 397 nm . a second diode laser at 866 nm is used as a repump .
both lasers are frequency locked to cavities using the pound - drever - hall method , and their frequencies can be varied by changing the cavity lengths with piezoelectric elements .
the intensity of the detection laser at 397 nm is adjusted to about 40@xmath9 and the intensity of the repump laser at 866 nm is adjusted to approximately 120@xmath9 .
the doppler cooling and repump lasers are overlapped and sent to the trap using a photonic crystal fiber .
the laser beam is aligned almost parallel to the surface of the trap and forms an angle of approximately @xmath10 with the @xmath11 and @xmath12 axes .
ion fluorescence is collected perpendicular to the trap plane using a lens system of na = 0.29 and split between a pmt and ccd camera on a 90:10 beam splitter .
in an ideal paul trap the ion is confined to a position at which the electric field due to the oscillating drive voltage on the rf electrodes is zero .
stray dc electric fields , however , push the ion off the rf node and the ion undergoes so - called micromotion driven by the oscillating rf field @xcite .
this motion causes broadening of the electronic transitions of the ion and , among other things , leads to a higher temperature limit for doppler cooling @xcite .
in addition , micromotion can lead to the heating of trapped ions due to the noise present at the secular sidebands of the micromotion drive @xcite . in order to position the ion on the rf node
, the dc potential is carefully adjusted to minimize micromotion .
crucial to all minimization schemes is the efficient detection of micromotion in all three spatial directions .
different techniques exist for fine - tuning the compensation of micromotion .
the photon correlation method@xcite relies on correlating the ion fluorescence to the phase of the rf field . in the resolved sideband method@xcite ,
the sidebands of a narrow atomic transition are compared with the carrier transition to estimate the modulation index .
both methods are widely used to suppress micromotion in 3d traps . however
, neither method can be easily extended to surface paul traps because they directly use the doppler shift induced by the ion motion . in the case of surface traps
, the geometry typically limits the laser alignment to be in the plane parallel to the trap surface resulting in no doppler shift associated with the oscillations perpendicular to the trap surface unless one directs laser light onto the trap electrodes .
however , it has been documented that uv light hitting the trap surface can lead to dramatic charge buildup even to the point where the trap becomes inoperable for days@xcite . to circumvent this obstacle , the infrared repump light in @xmath0ca@xmath1has
been used to detect the doppler shift perpendicular to the trapping plane for micromotion compensation @xcite . however , many ion species such as mg@xmath13 , al@xmath13 , hg@xmath13 , cd@xmath13 , be@xmath13 do not have such transitions in the infrared and other methods need to be employed . here
we compensate the micromotion perpendicular to the trap plane with the following method .
when the ion is displaced from the rf node , any voltage applied on the rf electrode creates an electric field at the ion position .
if this voltage contains a frequency component which is in resonance with one of the ion secular frequencies the ion can get excited in the direction of the secular mode provided that the driving field from the rf electrodes has some projection @xcite .
experimentally we find that large oscillation amplitudes of each of the three secular motions can be detected as a drop in ion fluorescence .
the dynamics of ion fluorescence in the presence of the cooling laser and a resonant excitation are complex and go beyond the scope of this study @xcite . minimizing micromotion
is achieved by shifting the ion position via dc potentials until the ion is in the rf minimum and can not be excited by driving the rf electrode at any of the secular frequencies .
this method is also being used by nist @xcite and osaska @xcite ion trap groups . in order to implement this method , we first position the ion within 1 @xmath2 m along the @xmath11 direction from the rf null using the ccd camera and varying the rf amplitude . further compensation in the @xmath11 direction
is achieved by reducing the linewidth of the s@xmath14-p@xmath14 transition . for this
, the detection laser intensity is adjusted close to saturation and red - detuned from the transition so that the fluorescence drops to half that of the value at the resonance .
then compensation voltages are adjusted to minimize the fluorescence .
both methods detect only micromotion along the _ x _ direction , i.e. the direction which is parallel to the trap surface .
for a very coarse compensation along the _ y _ direction ( perpendicular to the trap surface ) , we keep the frequency of 397 nm and 866 nm laser on resonance and maximize the ion fluorescence by adjusting the compensation voltages .
once a coarse compensation is achieved using the above methods , we proceed with the method as outlined in the beginning of this section . instead of exciting the secular frequency @xmath15 directly , we excite at a frequency @xmath16 @xcite . to achieve this ,
the excitation signal from a function generator is mixed with the trap drive @xmath17 before it is amplified and stepped up with a helical resonator , and scanned around @xmath18 .
when the frequency of the excitation becomes resonant with @xmath18 , the ion heats up resulting in a decrease in fluorescence ( see fig.[fig : mmdip ] ) , @xcite .
a crucial requirement is that the step - up circuit which produces the high - voltage trap drive has a large enough bandwidth .
the bandwidth of the helical resonator in our experiment is 270khz and allows compensation with excitation frequencies up to order @xmath192 mhz .
the frequency of the doppler cooling laser was detuned between 1 and 5 mhz below the @xmath20 transition to maximize sensitivity of the ion fluorescence to the ion kinetic energy .
+ for compensation of micromotion in the _ x _ direction , the excitation signal is scanned repeatedly around @xmath21 while adjusting the voltages on the dc electrodes such that between successive scans the static electric field minimum moves predominantly in the _ x _ direction .
the compensated position is reached when resonance of the excitation does not result in a decrease of fluorescence .
the same process is repeated for compensation in the _ y _ direction .
results are shown in fig.[fig : mmcomp ] .
fig.[fig : mmx2d ] and fig.[fig : mmy2d ] show the change in the dip depth when the excitation frequency is @xmath21 and @xmath22 , respectively , as a function of ion position along the _ x _ and the _ y _ direction .
the gray area in fig.[fig : mmx2d ] is the region were no data is acquired since the excitation would drive the ion out of the trap .
fig.[fig : mmx1d ] and fig.[fig : mmy1d ] show cross sections of the 2d plots along the equilibrium positions of the ion when the dc saddle point moves along the _ x _ and _ y _ directions , respectively .
the energy gain rate of the ion @xmath23 is expected to be @xmath24 , where @xmath25 is the strength of the exciting electric field .
we find that the data can be well fitted with a parabola suggesting that the dip depth is linear in the ion energy for our experiments while the electric field @xmath25 can be described well as a quadrupolar field .
the accuracy of our measurements is estimated by calibrating the ion displacement as a function of change in the dc voltages using the ccd camera in the x direction and the detection laser in the y direction .
these values were verified by modeling the displacement of the dc minimum with variation of the compensation voltages . by translating the applied voltage into actual displacement ,
we determine an accuracy of about 50 nm in the x direction and 300 nm in y direction in positioning the ion at the rf minimum .
this corresponds to excess micromotion amplitudes of 6 nm and 40 nm in these respective directions .
the accuracies could be further improved by increasing the excitation voltage and decreasing the frequency detuning of the detection laser from resonance .
one concern with this method is that in practical situations the dc electrodes pick rf voltage , the phase of which might be shifted , or which might depend on the rf excitation frequency .
the reason for this is that the dc electrodes may capacitively couple to the rf electrode , with a frequency dependent coupling determined by the filtering circuits connected to the dc electrodes . in our setup
, we estimate the rf pickup of the excitation at @xmath18 on the dc electrodes to differ by less than 20 @xmath2v from the pick up at @xmath17 which would shift the rf - null by about 15 pm , and thus not limit the accuracy .
it is instructive to extract the size and direction of the stray fields from the compensation voltages along various trapping positions .
we derive the stray field from the applied compensation voltages via an accurate electrostatic model of the trapping fields .
thus determining the stray fields over an extended region yields important information for shuttling experiments or investigating the mechanism for charge buildup .
we model the trap potentials using the boundary element method solver cpo @xcite .
since the radial trap axes are tilted , we determine the radial secular frequencies from a 2-dimensional polynomial fit in the @xmath7-plane .
the agreement between experiment and simulation is better than 4% , corresponding to a disagreement of less than 20khz for axial secular frequency of 500khz , and 30khz for radial frequencies of between 1 and 2mhz .
we attribute the disagreement to details of the trap electrode geometry which were not included in the electrostatic simulation , as well as to the unknown spatial inhomogeneity of the electrostatic stray fields .
only the spatially inhomogeneous part of the stray field will contribute to the discrepancy between simulation and experiment , since the electrostatic stray fields are compensated and included in the simulations , as we now describe .
the essential part of the stray field determination is to perform micromotion compensation as described above . after the compensation parameters
have been determined , we measure the ion position along the z direction using the ccd imaging system .
the values of the dc voltages at the compensated configuration and the ion position are then input to a minimization algorithm .
the algorithm finds the stray field which results in the observed compensation parameters and position .
the accuracy of our measurement scheme is limited by an estimated uncertainty of @xmath262.5 @xmath2 m in measuring the absolute ion position along the trap axis .
this position uncertainty arises from the imprecision of the alignment of the microscope objective used for imaging with respect to the trap plane and from the size of trap features used to determine the position along the trap axis .
this results in a systematic error of the electric field curves , mainly an offset of the entire curve .
we find typical inaccuracies of @xmath27v / m .
the precision in determining the electric field at each ion position is on the order of a few v / m , and is limited by the precision to which the compensation parameters and relative ion position are determined .
the errors from imperfect compensation lead to stray field imprecisions of @xmath260.4 v / m in the horizontal direction and @xmath262.5 v / m in the vertical direction , where we assume that the precision is limited by the ion displacement steps in fig.[fig : mmcomp ] .
the accuracy in determining the axial ion position at any given setting with respect to its positions in neighboring settings is limited by aberrations in the imaging optics and is estimated to be @xmath260.1@xmath2 m .
this limitation leads to an imprecision of @xmath28v / m in the electric fields .
the above limitations of the measurement leads to a total error of @xmath29v / m , however , the limitations are of technical nature and can be further improved .
stray electrostatic field along the trap axis .
the three components @xmath30 ( @xmath31 ) , @xmath32 ( @xmath33 ) , @xmath34 ( @xmath35 ) and the total electric field @xmath25 ( @xmath36 ) are shown .
the solid curves on the two sides of each curve correspond to the possible systematic offset arising from the uncertainty in the absolute axial position , as described in the text.,scaledwidth=50.0% ] [ sec : electric - stray - fields ] we performed this type of analysis over an extensive part of the trap .
measurements were performed during several months .
the results obtained in the first weeks of trap installation in vacuum and trap operation were significantly different than those obtained later on .
initially , the magnitude of the stray electric fields in the radial directions varied on a day - to - day basis with a mean of 47 v / m and a standard deviation of 52 v / m .
measurements during that time were carried out only for one isolated axial ion position at 2240 @xmath2 m .
after two months of trap operation , there was an abrupt change in stray electric fields to considerably higher values .
this change coincided with a pressure increase in the chamber to approximately @xmath37 mbar because of a temporary ion pump failure .
subsequently , the stray field was mapped along the trap axis over a length of approximately 2 mm .
the results are shown in fig.[fig : el - field ] .
we measured stray fields as large as 1300v / m .
intuitively it is conceivable that two localized charge sources located near the z axis are responsible for the observed electric field pattern .
however , such inverse problems can not be solved without further assumptions about the charge distribution @xcite .
attempts to reconstruct the charge distribution mathematically are in progress and will be reported elsewhere the position of the postulated charges close to the most frequently used trapping positions suggests that exposure of the trap material to laser light caused the long term charging .
finally , the high stray fields also appear to be spatially and temporally correlated with the observed high heating rate@xcite .
our measurement scheme allowed us to monitor charging of the trap during operation .
we found that the stray fields after the abrupt change in trap behavior are semi - permanent , showing a slow temporal drift on the order tens v / m per week .
the most striking occurrence of this slow drift is the slight discontinuity in the data of fig.[fig : el - field ] around an axial position of 2450@xmath38 .
this drift by 70v / m , corresponding to a fractional change in the stray field by @xmath396% , occurred over a period of 9 days .
besides the slow drift of the stray field , short - term charging is observed in the course of a day .
this is typically on the order of 100v / m .
when the laser light is turned off , discharging occurs over the course of several hours .
in this article , we demonstrate a simple , yet efficient method of measuring stray electric fields in planar ion traps .
this method permits us to sense electric fields over an extended region , thus providing insight into the undesired charging of ion traps .
this ability to characterize electric fields in the trapping region will be a valuable tool for evaluating planar ion traps , for developing ion loading approaches that minimize stray charging , and for compensating stray fields .
we expect that this technique will be useful not only for scalable quantum information processing , but also for precision frequency metrology applications of trapped ions .
the experiments were supported by the austrian ministry of sciences with a start grant and by the director , office of science , office of basic energy sciences , materials sciences and engineering division , of the u.s .
department of energy under contract no .
de - ac02 - 05ch11231 .
n. daniilidis was supported by the european union with a marie curie fellowship . f.
schmidt - kaler acknowledges support from the german - israel foundation and the eu network aqute .
10 t rosenband , d b hume , p o schmidt , c w chou , a a , l lorini , w h oskay , r e drullinger , t m fortier , j e stalnaker , s a a , w c swann , n r newbury , w m itano , d j wineland , and j c bergquist .
, 319(5871):18081812 , 2008 .
s seidelin , j chiaverini , r reichle , j j bollinger , d leibfried , j britton , j h wesenberg , r b blakestad , r j epstein , d b hume , w m itano , j d jost , c langer , r ozeri , n shiga , and d j wineland . , 96:253003 , 2006 .
j britton , d leibfried , j beall , r b blakestad , j j bollinger , j chiaverini , r j epstein , j d jost , d kielpinski , c langer , r ozeri , r reichle , s seidelin , n shiga , j h wesenberg , and d j wineland . . , 2006
d r leibrandt , j labaziewicz , r j clark ,
i l chuang , r epstein , c ospelkaus , j wesenberg , j bollinger , d leibfried , d wineland , d stick , j sterk , c monroe , c .- s .
pai , y low , r frahm , and r e slusher . . ,
9(11):0901 , 2009 . | we use a single ion as an movable electric field sensor with accuracies on the order of a few v / m . for this , we compensate undesired static electric fields in a planar rf trap and characterize the static fields over an extended region along the trap axis .
we observe a strong buildup of stray charges around the loading region on the trap resulting in an electric field of up to 1.3 kv / m at the ion position .
we also find that the profile of the stray field remains constant over a time span of a few months . |
the suppression of single inclusive high @xmath0 hadrons in heavy - ion collisions as compared to the scaled expectation from p - p collision has long been regarded as caused by energy loss of high @xmath2 partons into a dense partonic environment @xcite . at rhic ,
the nuclear suppression factor @xmath1 for pions in central au - au collisions has been measured out to 20 gev @xcite and a factor @xmath3 suppression observed .
the hope of using hard probes such as the single hadron suppression in the context of heavy - ion collisions is to infer properties of the medium and its density evolution from induced changes relative to the corresponding known hard process in p - p collisions .
there is , however , growing evidence that single hadron suppression is not enough to unambiguously determine even the distribution of energy / momentum shifts of partons traversing the medium @xcite .
however , if energy loss can not be determined reliably in momentum space , there is little hope to try to infer the qcd matter density distribution in position space .
back - to - back correlations of hard hadrons @xcite are a more differential hard process . due to the different geometrical averaging in the case of single hadron suppression vs. back - to - back suppression
, one may hope to obtain information about the spatial distribution of dense medium from combining the two observables .
while theoretical studies for back - to - back correlations as measured at rhic have been done @xcite , they seem to indicate that for rhic kinematics the amount of additional information is not substantial , in essence various models for the energy loss mechanism and for the density evolution which describe single hadron suppression also perform well when compared with dihadron suppression .
the notable exception is a class of energy loss models based on a purely linear dependence of the energy loss with pathlength those seem to be strongly disfavoured by back - to - back correaltions @xcite .
as suggested in @xcite , the reason why there is only little sensitivity to the qcd matter density distribution at rhic kinematics may be that the lever - arm in momentum is not large enough to probe substantial shifts in parton momentum for a steeply falling parton spectrum , even a moderate shift in parton momentum effectively resembles an absorption of partons , and this fact greatly reduces the sensitivity . at the lhc however where the partonic @xmath2 range is large this ceases to be a problem and consequently the suppression of hard back - to - back correlated hadrons becomes a promising probe . in this paper , we aim to provide a baseline prediction for the per - trigger yield in hard back - to - back correlations .
this complements a baseline prediction of the nuclear suppression factor @xmath1 @xcite made within the same model framework .
we compare with the calculation at rhic kinematics and point out similarities and differences .
as in @xcite we calculate the correlation strength of hadrons back to back with a hard trigger in a monte - carlo ( mc ) simulation .
there are three important building blocks to this computation : 1 ) the primary hard parton production , 2 ) the propagation of the partons through the medium and 3 ) the hadronization of the primary partons . only the step 2 )
probes medium properties , and hence it is here that we must specify details for the evolution of the qcd medium and for the parton - medium interaction . let us first discuss steps 1 ) and 3 ) which are common to the simulation in p - p and pb - pb collisions . in ref .
@xcite it has been demonstrated that leading order ( lo ) perturbative quantum chromodynamics ( pqcd ) is rather successful in describing the @xmath0-spectrum of inclusive hadron production over a wide range in @xmath4 when supplemented with a @xmath4-dependent @xmath5-factor to adjust the overall normalization .
this factor parametrizes next - to - leading order effects .
since we are in the following only interested in ratios of @xmath0-distributions , i.e. yields per triggered hadron , any factor independent of @xmath0 drops out .
hence , in the following we use lo pqcd expressions without trying to adjust the absolute normalization .
the production of two hard partons @xmath6 with transverse momentum @xmath2 in lo pqcd is described by @xmath7 where @xmath8 and @xmath9 stand for the colliding objects ( protons or nuclei ) and @xmath10 is the rapidity of parton @xmath11 .
the distribution function of a parton type @xmath12 in @xmath8 at a momentum fraction @xmath13 and a factorization scale @xmath14 is @xmath15 .
the distribution functions are different for the free protons @xcite and nucleons in nuclei @xcite .
the fractional momenta of the colliding partons @xmath12 , @xmath16 are given by @xmath17 + \exp[\pm y_2 ] \right)$ ] .
expressions for the pqcd subprocesses @xmath18 as a function of the parton mandelstam variables @xmath19 and @xmath20 can be found e.g. in @xcite . by selecting pairs of @xmath6
while summing over all allowed combinations of @xmath21 , i.e. @xmath22 where @xmath23 stands for any of the quark flavours @xmath24 we find the relative strength of different combinations of outgoing partons as a function of @xmath2 . for the present investigation , we require @xmath25 , i.e. we consider only back - to - back correlations detected at midrapidity . in the first step ,
we sample eq .
( [ e-2parton ] ) summed over all @xmath6 to generate @xmath2 for the event . in the second step
we perform a mc sampling of the decomposition of eq .
( [ e-2parton ] ) according to all possible combinations of outgoing partons @xmath6 at the @xmath2 obtained in the first step .
we thus end with a back - to - back parton pair with known parton types and flavours at transverse momentum @xmath2 . to account for various effects , including higher order pqcd radiation , transverse motion of partons in the nucleon ( nuclear ) wave function and effectively also the fact that hadronization is not a collinear process , we fold into the distribution @xmath26 an intrinsic transverse momentum @xmath27 with a gaussian distribution , thus creating a momentum imbalance between the two partons as @xmath28 .
we use a value of @xmath29 gev for rhic kinematics @xcite and observe a sensitivity of less than 10% of the away side yield to changes of this number by @xmath30% . while @xmath27 is expected to rise at lhc conditions , in the momentum region we wish to explore , i.e. @xmath31 gev there is no sensitivity of the calculation to the choice of intrinsic transverse momentum .
before we can count hadrons above certain trigger or associate - momentum threshold , we have to convert the simulated partons into hadrons .
more precisely , in order to determine whether there is a trigger hadron @xmath32 above a given threshold , given a parton @xmath33 with momentum @xmath2 , we need to sample @xmath34 , i.e. the probability distribution to find a hadron @xmath32 from the parton @xmath33 where @xmath32 is _ the most energetic hadron of the shower _ and carries the momentum @xmath35 . in previous works @xcite we have approximated this by the normalized fragmentation function @xmath36 , sampled with a lower cutoff @xmath37 which is adjusted to the reference d - au data .
this procedure can be justified by noting that only one hadron with @xmath38 can be produced in a shower , thus above @xmath39 the distributions @xmath40 and @xmath34 are ( up to the scale evolution ) identical , and only in the region of low @xmath41 where the fragmentation function describes the production of multiple hadrons do they differ significantly .
however , for a fixed cut in the hadronic momentum @xmath0 , low @xmath41 implies probing the high @xmath2 part of a steeply falling parton spectrum , thus this region is suppressed .
empirically , sampling @xmath40 has provided a good description of the baseline @xcite .
however , at the lhc energies , keeping the same lower cutoff @xmath37 is not acceptable as it would imply that there is no perturbative production of hadrons in the range below , say , 8 gev from a 100 gev parton .
presumably , a viable solution would be to readjust the cutoff to a measured set of reference data from p - p collisions . in the absence of such a data set ,
we have chosen a different path by extracting @xmath34 from shower simulations of quarks and gluons in pythia @xcite .
this moreover has an additional advantage , since there is a subtle point with the scale dependence of the hadronization functions . while @xmath40 is an empirical function where the scale is set by the produced hadron and the function
is folded with the full parton spectrum ( so that @xmath42 ) , @xmath34 is an object where the scale is set by the given parton energy and the outcome of the folding is a distribution of hadron momenta .
the latter is more appropriate for the situation realized in the mc simulation in which the partonic @xmath2 is fixed before hadronization . in fig .
[ f - ffcomp ] we show the extracted @xmath34 compared with @xmath36 from the kkp set of fragmentation functions @xcite in a probabilistic interpretation where we present the fragmentation function both at @xmath43 and @xmath44 , allowing a direct comparison at @xmath45 and @xmath46 . in the region @xmath47 , the quantities @xmath34 and @xmath36 in fig .
[ f - ffcomp ] differ from each other as expected . from the figure
, it is also evident that in the region @xmath48 where we expect agreement there remain differences between the empirical fragmentation function and the shower development in pythia which are beyond the uncertainties introduced by the qcd scale evolution . especially in the case of the gluon fragmentation
the differences are sizeable at higher @xmath41 .
we do not wish to enter a discussion of the physics origin of these differences at this point but rather accept @xmath49 from pythia as a trial ansatz which needs to be tested against data .
using @xmath34 on both outgoing partons from the primary pqcd vertex ( see eq .
[ e-2parton ] ) we find the two most energetic hadrons from both showers .
following the experimental analysis procedure , out of the two we define the harder one to be the trigger hadron if it passes a set momentum cut , otherwise we ignore the event .
the direction of the trigger hadron is taken to be the @xmath50 direction and the hemisphere corresponding to this direction is referred to as near side , the opposite hemisphere is then defined to be the away side. in order to compute the strength of hadrons correlated with a trigger on the away side , the leading contribution is found by considering the leading hadron produced from the away side parton for which @xmath51 is sufficient . however , in order to determine the correlation strength on the near side , the production of ( at least ) one more energetic hadron in the shower needs to be considered as the leading contribution here corresponds to the trigger hadron itself . thus , note that to the same level of approximation of the shower development the near side correlation is only found at next - to - leading ( nl ) fragmentation whereas an away side correlation appears already at leading fragmentation and receives a correction at next - to - leading fragmentation . in our framework
, we treat this next - to - leading contribution through the conditional probability density @xmath52 which for fixed @xmath53 corresponds to a probability density , i.e. @xmath54 .
this is conceptually similar to the dihadron fragmentation function introduced in @xcite .
the whole hadron production in the shower arises in this language as a tower of conditional probability denities @xmath55 with the probability to produce @xmath56 hadrons with momentum fractions @xmath57 from a parton with energy @xmath58 being @xmath59 .
in previous calculations @xcite we have modelled the next - to - leading conditional fragmentation probability using the experimentally measured probability distribution @xmath60 of associated hadron production in d - au collisions @xcite as a function of @xmath61 where @xmath61 is the fraction of the trigger hadron momentum carried by the associated hadron .
a factor @xmath62 on the near side and @xmath63 on the away side was included to enforce energy conservation . in this work ,
we use @xmath64 as extracted from pythia instead . by construction
, @xmath65 is satisfied , and energy - momentum conservation ( which the extracted distribution fulfills automatically ) implies @xmath66 , i.e. the region where @xmath52 is nonzero in the @xmath67 plane is a triangular region from the edges @xmath68 where @xmath69 to the center @xmath70 where @xmath71 .
the resulting distributions for u - quarks and gluons are shown in fig .
[ f - a2 ] .
we describe the production and spacetime evolution of qcd matter in central pb+pb collisions within the framework of pqcd + saturation + hydrodynamics ( the ekrt model ) @xcite . in this framework ,
primary parton production is computed by supplementing the collinearly factorized pqcd parton production of eq .
( [ e-2parton ] ) with the conjecture of saturation of produced gluons inhibiting the growth of gluon number at transverse momenta below the saturation scale , @xmath72 . in central heaviest - ion collisions at rhic and lhc
, the computation remains in the perturbative regime as @xmath73 gev @xmath74 .
the average formation time of the qcd system , @xmath75 fm/@xmath76 at rhic and lhc , and pqcd parton production in a central rapidity unit can then be computed , and thus the initial qgp densities estimated , see @xcite for details . with the assumptions of full thermalization and no initial transverse flow at @xmath77 , we then apply relativistic , longitudinally boost - symmetric , azimuthally symmetric , longitudinally and transversally expanding ideal hydrodynamics to describe the spacetime evolution of the formed qcd matter .
we use the results obtained in @xcite , where the qgp is assumed to be an ideal gas of gluons and three flavours of massless quarks and antiquarks , while the hadron resonance gas consists of all hadronic states up to mass 2 gev .
a first - order phase transition takes place at a critical temperature 165 mev , and , with binary - collision initial energy - density profiles , the system decouples at @xmath78 mev ( see also the discussion of the decoupling dynamics and @xmath79 in @xcite ) .
illustrations of the phase boundaries , flow lines and transverse velocity contours for central au+au collisinons at rhic and pb+pb at lhc , as well as extensive comparison with the hadron spectra measured at rhic and further discussion can be found in @xcite . within the framework
we apply here , there are obvious theoretical uncertainties in the extrapolation to the lhc energies , which are related to the gluon shadowing ( see @xcite ) in the primary production , and to the qcd matter equation of state and to the currently debated possible dissipative effects @xcite in the matter evolution itself . accepting these uncertainties
, however , the benefit in using the ekrt model is that being a closed framework it has predictive power : once @xmath79 has been fixed at rhic @xcite , straightforward rhic - tested predictions for the lhc hadron spectra can be made @xcite , and the evolution of qcd matter densities and flow velocities , the input for the present paper , obtained . in the case of heavy - ion collisions , partons emerging from a hard vertex
are surrounded by the soft bulk matter and may lose energy before hadronization .
we compute the energy loss of a hard parton traversing a thermalized medium using the formalism by baier , dokshitzer , muller , peigne and schiff ( bdmps ) @xcite which assumes that energy is predominantly lost due to medium - induced radiation .
this formalism is cast into the form of energy loss probability distributions , so - called quenching weights in @xcite .
this is a convenient formulation for the purpose of a mc simulation .
the key quantity characterizing the energy loss induced by a medium with energy density @xmath80 in the bdmps formalism @xcite is the local transport coefficient @xmath81 which characterizes the squared average momentum transfer from the medium to the hard parton per unit pathlength .
since we consider a time - dependent inhomogeneous medium , this quantity depends on proper time @xmath82 , spacetime rapidity @xmath83 , cylindrical radius @xmath84 and in principle also on azimuthal angle @xmath85 , but for the time being we focus on central collisions only .
the transport coefficient is related to the energy density of the medium @xmath80 as @xmath86^{3/4 } ( \cosh \rho - \sinh \rho \cos\alpha)\ ] ] with @xmath87 for an ideal quark - gluon plasma ( qgp ) @xcite here , @xmath88 is the local flow rapidity with angle @xmath89 between flow and parton trajectory @xcite .
the parameters @xmath80 and @xmath88 characterize the medium and must be supplied by the underlying qcd medium evolution model . in the following , motivated by ref . @xcite we assume the proportionality constant @xmath5 to remain unaltered in different phases of the medium . we treat @xmath5 as an adjustable parameter ( for a detailed discussion see @xcite ) which however , once fixed , is independent on @xmath4 of the heavy - ion collision and assumes the same value at rhic and lhc energies while the medium density given by @xmath90 changes substantially . given the local transport coefficient at each spacetime point , a parton s energy loss depends on the position of the hard vertex at @xmath91 in the transverse plane at @xmath92 and the angular orientation of its trajectory @xmath85 ( i.e. its path through the medium ) . in order to determine the probability @xmath93 for a hard parton with energy @xmath94 to lose the energy @xmath95 while traversing the medium on its trajectory , we make use of a scaling law @xcite which allows to relate the dynamical scenario to a static equivalent one by calculating the following quantities averaged over the jet trajectory @xmath96 @xmath97 and @xmath98 as a function of the jet production vertex @xmath99 and its angular orientation @xmath85 .
we set @xmath100 whenever the decoupling temperature of the medium @xmath101 is reached . using the numerical results of @xcite
, we obtain @xmath102 for @xmath103 and @xmath104 as a function of jet production vertex and the angle @xmath85 corresponding to @xmath105 \delta\left ( \delta e - \sum_{i=1}^n \omega_i\right ) \exp\left[-\int d\omega\frac{di}{d\omega } \right]\ ] ] which makes use of the distribution @xmath106 of gluons emitted into the jet cone .
the explicit expression of this quantity for the case of multiple soft scattering can be found in @xcite .
note that the formalism of @xcite is defined for the limit of asymptotic parton energy , hence the probability distribution obtained @xmath102 is independent of @xmath94 . the initially produced hard parton spectrum and ,
consequently , the number of hard vertices in the @xmath107 plane ( where the @xmath41-axis is given by the beam direction ) in an @xmath108 collision at fixed impact parameter * b * , are proportional to the nuclear overlap , @xmath109 where @xmath110 is the standard nuclear overlap function .
we define a normalized geometrical distribution @xmath111 for central collisions as @xmath112 ^ 2}{t_{aa}(0)},\ ] ] where the thickness function is given in terms of the nuclear density @xmath113 as @xmath114 . for a given event in the mc simulation , we first sample eq .
( [ e - geo ] ) to determine the vertex position in the transverse plane . from this point , we propagate the parton through the medium and we evaluate eqs .
( [ e - omega],[e - ql ] ) along the path , on which the medium properties enter via eq .
( [ e - qhat ] ) . computing the corresponding @xmath102 for each parton , we sample this distribution to determine the actual energy loss for near and away side parton in the event . in the case @xmath115 we consider the parton to be absorbed by the medium and not contributing to high @xmath2 yield .
the first step is to test the mc simulation without a medium to see how well lo pqcd combined with the first two terms in the hadronization probability distribution expansion , i.e. @xmath49 and @xmath52 describes the data . for this purpose ,
we compare in fig .
[ f - baseline ] with the correlation strength measured in d - au collision by the star collaboration @xcite .
the data have been taken for a trigger hadron between 8 and 15 gev and in two associate yield bins ( on both near and away side ) of 4 - 6 and 6 + gev . in order to simulate d - au collisions , we insert a nuclear parton distribution function into eq .
( [ e-2parton ] ) and assume that no medium is formed and that nuclear matter is not dense enough to induce significant energy loss . as apparent from the figure ,
the calculation is able to describe the data generally rather well , however in the 4 - 6 gev momentum bin on the away side it overshoots the measured data .
we have re - run the simulation taking into account only the leading fragmentation term @xmath49 and find that this contribution alone seems to do well with the data . since the trigger is always chosen to be the most energetic hadron in the simulation
, there is considerable bias towards the trigger parton being a quark , as the quark fragmentation is harder ( cf .
[ f - ffcomp ] ) . in the kinematical range probed , the @xmath116 subprocess is rather important , thus there is a bias towards a gluon being the away side parton .
we note from fig .
[ f - ffcomp ] that the discrepancy between the gluon fragmentation function and @xmath117 is larger than for quarks .
quite likely , this discrepancy also propagates into the extraction of @xmath118 from pythia and is the reason that the present calculation shows differences here relative to the data ( and to previous results @xcite ) .
the eventual solution to this discrepancy would be to find a different way to extract @xmath49 and @xmath52 than pythia shower simulations .
however , since the overall description of the data seems to be reasonable , we proceed with the hadronization procedure as introduced above , but assign an uncertainty to the simulations of the heavy - ion case which we estimate by calculating the results with and without @xmath52 on the away side . in the next step ,
we test the mc simulation including in - medium energy loss with a soft medium background evolution as provided by a hydrodynamical calculation for 200 agev central au - au collisions .
the parameter @xmath5 in eq .
( [ e - qhat ] ) is fixed by the single hadron suppression for this system @xcite , thus the computation proceeds without any further free parameters .
the resulting yields per trigger as a function of associate hadron momentum for near and away side are shown in fig .
[ f - auau ] .
we observe that the yield on the near side is almost identical to the calculated yield in d - au collisions and systematically below the data
. one way of resolving the discrepancy while retaining agreement with the d - au data could be the inclusion of nnl fragmentation processes . due to the large errors in the d - au measurement
, this would not destroy the agreement with the baseline . while a full computation of the nnl conditional fragmentation probability distribution @xmath119 is very involved , we have checked that the approximation @xmath120 produces the right order of magnitude to agree with the data .
a different possibility is a sizeable contribution of coalescence processes @xcite in the au - au case to the near side hadron yield which is not included in the calculation .
if coalescence processes are relevant , one would expect to systematically underestimate the data . given the large uncertainty of the baseline d - au measurement and the theoretical uncertainty in the shower modelling using pythia
, it is unfortunately impossible to make a definite statement what the relevant physics process underlying the discrepancy is . on the away side ,
the calculation shows good agreement in both the 4 - 6 gev and the 6 + gev momentum bin .
however , given that the calculation overshoots the d - au baseline in the lower momentum bin , one has to conclude that the agreement in this bin is accidental .
this is clearly confirmed by omitting nl fragmentation on the away side ( which restores agreement with the baseline ) : doing so results in an underprediction of this bin just as observed in our previous calculation @xcite which is consistent with the observation that coalescence processes @xcite play a substantial role for hadronization in this momentum region . in fig .
[ f - z_t ] , we show the distribution of @xmath121 on the away side in d - au and central au - au collisions and compare with data .
we observe a general tendency that the calculation slightly underestimates the data in the high @xmath61 region .
this is consistent with the observation that there is a trend that the calculation may underestimate the higher associate momentum window slightly .
however , within errors , the results are compatible .
we compute the expected correlation strength pattern for an energy of 5.5 atev for p - p collisions and pb - pb collisions as expected to take place at the cern lhc . for the trigger range
, we assume 50 - 70 gev to probe a regime which can be accessed at lhc but is beyond the reach of the rhic experiment . since our formalism includes shower development only to the second term @xmath52 , the associate hadron momentum cuts can not be too low , otherwise the main contribution to hadron production can not be assumed to be dominated by the leading and next to leading hadron .
thus , we start considering associate hadron production for 20 gev and above up to the trigger energy in 5 gev momentum windows . in order to account for the uncertainty with regard to the nl fragmentation on the away side
as discussed earlier , we carry out the simulation both including and excluding this contribution to provide an error estimate . in fig .
[ f - pbpb ] we show the resulting yield per trigger on the near and away side .
let us first focus on the away side ( right panel ) .
regardless if one uses nl fragmentation contributions or not , there is a suppression of the away - side yield when going from p - p to pb - pb collisions of about a factor 3 .
this is comparable with the factor 4 suppression observed at rhic .
the similarity is not quite unexpected , since in the range of the trigger momentum the same model predicts a single hadron suppression factor of about the magnitude seen at rhic @xcite .
the underlying reason is that the ratio between the typical scale of energy loss vs. the momentum scale probed is similar and the change in characteristic length scales between central au - au and central pb - pb is negligible .
up to a momentum scale of about 35 gev , there is an uncertainty associated with nl fragmentation , for higher momenta the uncertainty goes away .
let us now focus on the near side ( left panel in fig .
[ f - pbpb ] ) : here we observe that the yield per trigger is enhanced by 10 - 20% in pb - pb relative to to p - p collisions . at first glance , this seems to be surprising .
however , one should keep in mind that the model still includes a strong suppression of trigger hadrons , only the per - trigger yield is enhanced .
these findings can be reconciled by the observation that the dense medium at lhc acts like a gluon filter and in essence enhances the relative abundance of quarks escaping the medium to gluons escaping the medium .
the underlying reason is that gluons , due to their different color charge , couple more strongly to the medium than quarks and consequently lose more energy .
however , due to the harder leading and nl fragmentation , quark jets on average produce harder associated yield than gluon jets , which is precisely what is seen in the simulation . finally , in fig . [ f - pshift ] we take a closer look at the momentum spectrum of hadrons recoiling from the trigger . to first order ,
one finds the reduction by a factor three mentioned above when going from p - p to pb - pb collisions .
this would be consistent with complete absorption of 2/3 of all partons whereas 1/3 passes the medium without any energy loss .
however , as a closer comparison reveals , there is a momentum dependent distortion of the spectra in heavy - ion collisions which is associated with events in which an away - side parton emerges from the medium after a finite energy loss @xmath95 . from this @xmath2
dependent shift of the spectrum , one can in principle derive constraints about the averaged energy loss probability @xmath122 and hence about the medium density distribution .
we have made a baseline prediction for the expected per - trigger yield of associate hadrons as a function of transverse momentum in central pb - pb collisions at 5.5 atev for a trigger energy of 50 - 70 gev .
we have demonstrated that the effect of the medium is expected to leave a characteristic imprint in the observables . on the near side
, we expect enhancement of the associate hadron yield as compared to the p - p case due to the fact that energy loss of gluons is stronger , hence there is some bias towards quarks leading to a trigger hadron , and the fact that the nl fragmentation for quarks is harder leads to a slight enhancement in the relevant momentum region . on the away side
, we expect a substantial suppression of the yield as compared to the p - p case , although the suppression we find is in relative terms smaller than what is currently observed at rhic .
we have further shown that the shape of the momentum spectrum of away - side hadrons is not equal to the momentum spectrum expected for p - p collisions .
this is because the medium does not only act by absorbing partons , but one gets a chance to observe the characteristic shift in parton energy induced by the medium .
this allows a more detailed view on the geometry - averaged energy loss probability distribution . in this respect ,
the observable shows a potential which may not be accessible at rhic kinematics .
there are multiple uncertainties associated with this prediction at each step of the mc simulation . in the lo pqcd calculation
, both nlo corrections and the choice of the nuclear parton distribution functions result in an uncertainty , especially as the role of gluon shadowing can not be reliably extrapolated to the lhc regime .
however , we have tested that the difference between the npdf @xcite and the eks98 @xcite nuclear parton distribution sets is smaller than the statistical errors of the mc simulation . while the yield of partons at given @xmath2 is sensitive to the nuclear parton distribution ,
the associated per - trigger yield is not since many uncertainties cancel in the ratio .
there is likewise an uncertainty associated with the choice of the hadronization scheme .
we have tried to show the magnitude of this uncertainty by comparing with available rhic data and by indicating the relative magnitude of the leading and next - to - leading fragmentation term in the prediction .
a large uncertainty must also be assigned to the hydrodynamical prediction of the underlying medium evolution . at present , neither entropy production and the distribution in the initial state nor the equilibration process or the equation of state governing a hydrodynamical expansion are precisely known .
yet all these quantities would leave a characteristic signature on the geometrical bias induced by the requirement to obtain a trigger . keeping in mind these uncertainties ,
the prediction presented here should in any case show the right order of magnitude of the correlation strength .
furthermore , we believe the essential effects are described qualitatively correct , given that the radiative energy loss picture is valid . | the suppression of high transverse momentum ( @xmath0 ) inclusive hadron spectra in heavy - ion collisions as compared to the scaled expectation of high @xmath0 hadron production in p - p collisions is the most direct manifestation of the interaction of hard partons with the soft bulk medium produced in heavy - ion collisions which is absent in p - p collisions . yet
the measured nuclear suppression factor @xmath1 is a very averaged quantity and hence only a limited amount of information about the medium evolution and the nature of the interaction with the medium can be deduced from @xmath1 .
measurements of hard back - to - back hadron correlations in 200 agev au - au collisions at rhic have opened a new window to study the energy loss of partons in a medium in a more differential way and for a different distribution of in - medium pathlengths than in the case of @xmath1 . in this work
, we present an extrapolation of our results for back - to - back yields at rhic energies to 5.5 tev pb - pb collisions at the cern lhc .
we also discuss differences and similarities between the measurement at rhic . |
let us now focus on strategies to extract @xmath0 from pairs of @xmath3-meson decays , which are related to each other through the @xmath2-spin flavor symmetry of strong interactions . in order to deal with non - leptonic
@xmath3 decays , @xmath2-spin offers an important tool , and first approaches to extract ckm phases were already pointed out in 1993 @xcite .
however , the great power of the @xmath2-spin symmtery to determine weak phases and hadronic parameters was noticed just recently in the strategies proposed in @xcite@xcite .
since these methods involve also decays of @xmath34-mesons , @xmath3 experiments at hadron colliders are required to implement them in practice . at tevatron - ii , we will have first access to the corresponding modes and interesting results are expected @xcite . in the era of btev and the lhc , the @xmath2-spin strategies can then be fully exploited @xcite , as emphasized by stone at this symposium . in the following discussion , we shall focus on two particularly promising approaches , using the @xmath35 , @xmath36 @xcite and @xmath37 , @xmath38 @xcite systems .
looking at the corresponding feynman diagrams , we observe that @xmath36 is obtained from @xmath40 by interchanging all down and strange quarks .
the structure of the corresponding decay amplitudes is given as follows @xcite : @xmath41\ ] ] @xmath42,\ ] ] where @xmath43 , @xmath44 are cp - conserving strong amplitudes , and @xmath45 , @xmath46 measure , sloppily speaking , ratios of penguin to tree amplitudes . using these general parametrizations ,
we obtain the following expressions for the direct and mixing - induced cp asymmetries : @xmath47 @xmath48 consequently , we have four observables , depending on six `` unknowns '' .
however , since @xmath40 and @xmath36 are related to each other by interchanging all down and strange quarks , the @xmath2-spin flavor symmetry of strong interactions implies @xmath49 using this relation , the four observables ( [ asym-1])([asym-4 ] ) depend on the four quantities @xmath50 , @xmath51 , @xmath52 and @xmath0 , which can hence be determined .
it should be emphasized that no dynamical assumptions about rescattering processes have to be made in this approach , which is an important conceptual advantage in comparison with the @xmath1 strategies discussed above .
the theoretical accuracy is hence only limited by @xmath2-spin - breaking effects .
theoretical considerations allow us to gain confidence into ( [ u - spin - rel ] ) , which does not receive @xmath2-spin - breaking corrections in factorization @xcite . moreover , there are general relations between observables of @xmath2-spin - related decays , allowing experimental insights into @xmath2-spin breaking @xcite
. the @xmath2-spin arguments can be minimized , if we employ the @xmath53@xmath54 mixing phase @xmath52 as an input , which can be determined straightforwardly through @xmath55 .
the observables @xmath56 and @xmath57 allow us then to eliminate the strong phase @xmath51 and to determine @xmath50 as a function of @xmath0 .
analogously , @xmath58 and @xmath59 allow us to eliminate the strong phase @xmath60 and to determine @xmath61 as a function of @xmath0 .
the corresponding contours in the @xmath0@xmath50 and @xmath0@xmath61 planes can be fixed in a _ theoretically clean _ way .
using now the @xmath2-spin relation @xmath62 , these contours allow the determination both of the ckm angle @xmath0 and of the hadronic quantities @xmath50 , @xmath51 , @xmath60 ; for a detailed illustration , see @xcite .
this approach is very promising for tevatron - ii and the lhc era , where experimental accuracies for @xmath0 of @xmath63 @xcite and @xmath64 @xcite may be achieved , respectively .
it should be emphasized that not only @xmath0 , but also the hadronic parameters @xmath50 , @xmath51 , @xmath60 are of particular interest , as they can be compared with theoretical predictions , thereby allowing valuable insights into hadron dynamics . for strategies to probe @xmath0 and constrain hadronic penguin parameters using a variant of the @xmath35 , @xmath36 approach , where the latter decay is replaced through @xmath37 , the reader is referred to @xcite .
another interesting pair of decays , which are related to each other by interchanging all down and strange quarks , is the @xmath66 , @xmath67 system @xcite . in the strict @xmath2-spin limit
, the corresponding decay amplitudes can be parametrized as follows : @xmath68 @xmath69 where @xmath70 denotes a cp - conserving complex amplitude , @xmath71 , @xmath72 is a real parameter , and @xmath73 a cp - conserving strong phase . at first sight , it appears as if @xmath0 , @xmath72 and @xmath73 could be determined from the ratio of the cp - averaged rates and the two cp asymmetries provided by these modes .. ] however , because of the relation @xmath74 latexmath:[\[\label{ru - rel } = -\left[|a(b^0_s\to\pi^+k^-)|^2-|a(\overline{b^0_s}\to\pi^-k^+ ) that the three parameters @xmath0 , @xmath72 and @xmath73 can not be determined . to this end
, the overall normalization @xmath70 has to be fixed , requiring a further input . assuming that rescattering processes play a minor rle and that color - suppressed ew penguins can be neglected as well
, the isospin symmetry implies @xmath76 in order to extract @xmath0 and the hadronic parameters , it is useful to introduce observables @xmath77 and @xmath78 by replacing @xmath37 through @xmath38 in ( [ mixed - obs ] ) . using ( [ par1 ] ) , ( [ par2 ] ) and ( [ par3 ] ) , we then obtain @xmath79 @xmath80 together with the parametrization for @xmath81 as sketched in ( [ obs - par ] ) , these observables allow the determination of all relevant parameters .
the extraction of @xmath0 and @xmath73 is analogous to the `` mixed '' @xmath37 , @xmath82 approach discussed above .
however , now the advantage is that the @xmath2-spin counterparts @xmath38 of @xmath37 allow us to determine also the parameter @xmath72 without using arguments related to factorization @xcite : @xmath83}.\ ] ] on the other hand , we still have to make dynamical assumptions concerning rescattering and color - suppressed ew penguin effects .
the theoretical accuracy is further limited by @xmath84-breaking effects .
an interesting consistency check is provided by the relation @xmath85 , which is due to ( [ ru - rel ] ) .
a variant of this approach using the ckm angle @xmath39 as an additional input was proposed in @xcite .
there are many strategies to determine @xmath0 , suffering unfortunately in several cases from experimental problems .
the approaches discussed above , employing penguin processes , are on the other hand very promising from a practical point of view and exhibit further interesting features . as a by - product , they also allow us to determine strong phases and other hadronic parameters , allowing comparisons with theoretical predictions . moreover , these strategies are sensitive probes for the physics lying beyond the standard model , which may lead to discrepancies in the extraction of @xmath0 or the hadronic quantities .
let us hope that signals for new physics will actually emerge this way .
b. aubert _ et al . _ [ babar collaboration ] , phys .
lett . * 87 * , 091801 ( 2001 ) [ hep - ex/0107013 ] ; + k. abe _ et al . _ [ belle collaboration ] , phys . rev .
lett . * 87 * , 091802 ( 2001 ) [ hep - ex/0107061 ] . for recent reviews ,
see y. nir , wis-18 - 01-aug - dpp [ hep - ph/0109090 ] ; + a. j. buras , tum - hep-402 - 01 [ hep - ph/0101336 ] ; + j. l. rosner , efi-2000 - 47 [ hep - ph/0011355 ] ; + r. fleischer , desy-00 - 170 [ hep - ph/0011323 ] . for recent analyses ,
see a. hcker _ et al .
_ , eur .
j. c * 21 * , 225 ( 2001 ) [ hep - ph/0104062 ] ; + m. ciuchini _ et al .
_ , jhep * 0107 * , 013 ( 2001 ) [ hep - ph/0012308 ] ; + a. ali and d. london , eur .
j. c * 18 * , 665 ( 2001 ) [ hep - ph/0012155 ] .
m. beneke _ et al .
_ , phys . rev
. lett . * 83 * , 1914 ( 1999 ) [ hep - ph/9905312 ] .
m. beneke _ et al .
_ , nucl .
b * 606 * , 245 ( 2001 ) [ hep - ph/0104110 ] .
li and h.l .
yu , phys .
d * 53 * , 2480 ( 1996 ) [ hep - ph/9411308 ] ; + y.y .
keum , h .-
li and a. i. sanda , phys .
b * 504 * , 6 ( 2001 ) [ hep - ph/0004004 ] ; + phys . rev .
d * 63 * , 054008 ( 2001 ) [ hep - ph/0004173 ] ; + y.y .
keum and h .- n .
li , phys .
d * 63 * , 074006 ( 2001 ) [ hep - ph/0006001 ] .
sachs , efi-85 - 22 ( unpublished ) ; + i. dunietz and r. g. sachs , phys .
d * 37 * , 3186 ( 1988 ) [ e : phys .
d * 39 * , 3515 ( 1988 ) ] ; + i. dunietz , phys .
b * 427 * , 179 ( 1998 ) [ hep - ph/9712401 ] ; + m. diehl and g. hiller , phys .
b * 517 * , 125 ( 2001 ) [ hep - ph/0105213 ] .
r. aleksan , i. dunietz and b. kayser , z. phys .
c * 54 * , 653 ( 1992 ) ; + r. fleischer and i. dunietz , phys .
b * 387 * , 361 ( 1996 ) [ hep - ph/9605221 ] ; + a. f. falk and a. a. petrov , phys .
* 85 * , 252 ( 2000 ) [ hep - ph/0003321 ] ; + d. london , n. sinha and r. sinha , phys .
lett . * 85 * , 1807 ( 2000 ) [ hep - ph/0005248 ] . m. gronau and d. wyler , phys .
b * 265 * , 172 ( 1991 ) ; + d. atwood , i. dunietz and a. soni , phys .
lett . * 78 * , 3257 ( 1997 ) [ hep - ph/9612433 ] ; + m. gronau , phys .
d * 58 * , 037301 ( 1998 ) [ hep - ph/9802315 ] .
r. fleischer and d. wyler , phys .
d * 62 * , 057503 ( 2000 ) [ hep - ph/0004010 ]
. m. gronau , j. l. rosner and d. london , phys .
lett . * 73 * , 21 ( 1994 ) [ hep - ph/9404282 ] .
o. f. hernndez _ et al .
_ , phys .
b * 333 * , 500 ( 1994 ) [ hep - ph/9404281 ] ; + m. gronau _ et al .
_ , phys .
d * 50 * , 4529 ( 1994 ) ; phys .
d * 52 * , 6356 ( 1995 ) [ hep - ph/9504326 ] .
r. fleischer , phys .
b * 365 * , 399 ( 1996 ) [ hep - ph/9509204 ] .
r. fleischer and t. mannel , phys .
d * 57 * , 2752 ( 1998 ) [ hep - ph/9704423 ] .
m. gronau and j. l. rosner , phys .
d * 57 * , 6843 ( 1998 ) [ hep - ph/9711246 ] .
r. fleischer , eur .
j. c * 6 * , 451 ( 1999 ) [ hep - ph/9802433 ] .
m. neubert and j. l. rosner , phys .
b * 441 * , 403 ( 1998 ) [ hep - ph/9808493 ] ; + phys .
* 81 * , 5076 ( 1998 ) [ hep - ph/9809311 ] .
m. neubert , jhep * 9902 * , 014 ( 1999 ) [ hep - ph/9812396 ] .
a. j. buras and r. fleischer , eur .
j. c * 11 * , 93 ( 1999 ) [ hep - ph/9810260 ] .
a. j. buras and r. fleischer , eur .
j. c * 16 * , 97 ( 2000 ) [ hep - ph/0003323 ] .
l. wolfenstein , phys .
d * 52 * , 537 ( 1995 ) ; + j. f. donoghue _ et al .
_ , phys .
lett . * 77 * , 2178 ( 1996 ) [ hep - ph/9604283 ] ; + m. neubert , phys .
b * 424 * , 152 ( 1998 ) [ hep - ph/9712224 ] ; + j. m. grard and j. weyers , eur .
j. c * 7 * , 1 ( 1999 ) [ hep - ph/9711469 ] ; + a. j. buras , r. fleischer and t. mannel , nucl .
b * 533 * , 3 ( 1998 ) [ hep - ph/9711262 ] ; + a. f. falk , a. l. kagan , y. nir and a. a. petrov , phys .
d * 57 * , 4290 ( 1998 ) [ hep - ph/9712225 ] ; + r. fleischer , phys .
b * 435 * , 221 ( 1998 ) [ hep - ph/9804319 ] .
d. cronin - hennessy _ et al .
_ [ cleo collaboration ] , phys .
lett . * 85 * , 515 ( 2000 ) .
b. aubert _ et al . _ [ babar collaboration ] , phys .
lett . * 87 * , 151802 ( 2001 ) [ hep - ex/0105061 ] .
k. abe _ et al . _ [ belle collaboration ] , phys .
lett . * 87 * , 101801 ( 2001 ) [ hep - ex/0104030 ] .
r. fleischer and t. mannel , ttp-97 - 22 , [ hep - ph/9706261 ] ; + d. choudhury , b. dutta and a. kundu , phys .
b * 456 * , 185 ( 1999 ) [ hep - ph/9812209 ] ; + x. he , c. hsueh and j. shi , phys .
lett . * 84 * , 18 ( 2000 ) [ hep - ph/9905296 ] ; + r. fleischer and j. matias , phys .
d * 61 * , 074004 ( 2000 ) [ hep - ph/9906274 ] ; + y. grossman , m. neubert and a. l. kagan , jhep * 9910 * , 029 ( 1999 ) [ hep - ph/9909297 ] ; + j. matias , uab - ft-514 [ hep - ph/0105103 ] .
for sophisticated fits , see h. lacker , these proceedings , and + http://www.slac.stanford.edu/@xmath86laplace/ckmfitter.html .
r. fleischer , eur .
j. c * 10 * , 299 ( 1999 ) [ hep - ph/9903455 ] .
r. fleischer , phys .
b * 459 * , 306 ( 1999 ) [ hep - ph/9903456 ] .
m. gronau and j. l. rosner , phys .
b * 482 * , 71 ( 2000 ) [ hep - ph/0003119 ] .
z. skands , jhep * 0101 * , 008 ( 2001 ) [ hep - ph/0010115 ]
. m. tanaka [ cdf collaboration ] , nucl .
instrum .
a * 462 * , 165 ( 2001 ) .
p. ball _ et al .
_ , cern - th-2000 - 101 [ hep - ph/0003238 ] .
r. fleischer , phys .
d * 60 * , 073008 ( 1999 ) [ hep - ph/9903540 ] .
m. gronau , phys .
b * 492 * , 297 ( 2000 ) [ hep - ph/0008292 ] .
r. fleischer , eur .
j. c * 16 * , 87 ( 2000 ) [ hep - ph/0001253 ] . c. chiang and l. wolfenstein , phys .
b * 493 * , 73 ( 2000 ) [ hep - ph/0004255 ] . | after a brief look at the well - known standard approaches to determine the angle @xmath0 of the unitarity triangle , we focus on two kinds of strategies , employing @xmath1 modes , and @xmath2-spin - related @xmath3 decays .
interesting `` puzzles '' , which may already be indicated by the present @xmath3-factory data , are pointed out , and the importance of the extraction of hadronic parameters , which are provided by these strategies as by - products , is emphasized .
[ cols= " < " , ] in fig . [ fig1 ] , we show the dependence of ( [ const1 ] ) and ( [ const2 ] ) on @xmath0 for the neutral @xmath1 system ; the charged @xmath1 curves look very similar @xcite . here
the crossed region below the @xmath4 curve , which is described by ( [ const2 ] ) , is excluded . on the other hand ,
the shaded region is the allowed range ( [ const1 ] ) for @xmath5 , arising in the case of @xmath6 .
this figure allows us to read off immediately the allowed range for @xmath0 for a given value of @xmath5 . in order to illustrate this feature ,
let us assume that @xmath7 has been measured , which would be in accordance with the central values of cleo and belle in table [ tab1 ] . in this case , the @xmath4 curve implies @xmath8 .
if we use additional information on @xmath9 , we may put even stronger constraints on @xmath0 . for @xmath6 ,
we obtain , for instance , the allowed range @xmath10 .
it is interesting to note that the @xmath4 curve is only effective for @xmath11 .
assuming @xmath12 to illustrate implications of the cp - averaged charged @xmath1 branching ratios , ( [ const2 ] ) is not effective and @xmath13 has to be fixed in order to constrain @xmath0 .
using @xmath14 , we obtain @xmath15 .
although it is too early to draw definite conclusions , it is important to emphasize that the second quadrant for @xmath0 , i.e. @xmath16 , would be preferred in these cases .
interestingly , such a situation would be in conflict with the standard analysis of the unitarity triangle @xcite , yielding @xmath17 . here
the stringent present experimental lower bound on @xmath18 implies @xmath19 .
another `` puzzle '' may arise from cp - conserving strong phases , which can also be constrained through the observables @xmath20 @xcite .
interestingly , the cleo and belle data point towards @xmath21 and @xmath22 , whereas we would expect equal signs for these quantities .
moreover , @xmath22 would be in conflict with factorization . consequently , the present cleo and belle data point towards a `` puzzling '' situation , whereas no such discrepancies are indicated by the results of the babar collaboration .
it is of course too early to draw any definite conclusions .
however , if future data should confirm such a picture , it may be an indication for new physics or large flavor - symmetry - breaking effects @xcite .
further studies are desirable to distinguish between these cases .
since @xmath1 modes are governed by penguin processes , they actually represent sensitive probes for new physics @xcite . due to
the recent theoretical progress in the description of @xmath23 decays , the theoretical uncertainties of @xmath24 and @xmath25 can be reduced to the level of @xcite @xmath26 and confidence into dynamical assumptions related to rescattering effects can be gained .
making more extensive use of qcd factorization , approaches complementary to the ones discussed above , which rely on a `` minimal '' input from theory , are provided . as a first step , we may use that the cp - conserving strong phase @xmath27 is predicted to be very small in qcd factorization , so that @xmath28 governing @xmath29 is close to one . as a second step ,
information on @xmath0 can be obtained from the predictions for the branching ratios and the observables @xmath30 .
finally , the information from all cp - averaged @xmath23 branching ratios can be combined into a single global fit for the allowed region in the @xmath31@xmath32 plane @xcite . for these approaches , it is of course crucial that the corrections entering in the qcd factorization formalism at the @xmath33 level can be controlled reliably . as argued in a recent paper @xcite , non - perturbative contributions with the same quantum numbers as penguin topologies with internal charm- and up - quark exchanges may play an important rle in this context .
the issue of @xmath33 corrections in phenomenological analyses will certainly continue to be a hot topic in the future . |
the system kcuf@xmath0 has long been known as a paradigm for an orbitally ordered system where a cooperative jahn - teller ( jt ) distortion is strongly competing with the electronic degrees of freedom as the driving force behind the orbital order.@xcite this system was investigated recently by realistic band structure calculations as a benchmark system for modeling structural relaxation effects due to electronic correlations @xcite and for revealing the influence of electronic superexchange on the orbital ordering.@xcite the compound seems to be orbitally ordered throughout its solid phase , but shows long - range a - type antiferromagnetic ( afm ) ordering only below @xmath8 = 39 k. in literature an orbital ordering temperature of about 800 k is often evoked in this system , but astonishingly experimental evidence for a transition at this temperature seems to be evasive .
early on , however , it was reported that between 670 k and 720 k an irreversible transition takes place.@xcite recently , the melting of the cooperative jt - transition has been studied in kcu@xmath9mg@xmath10f@xmath0 and from the extrapolation to undoped kcuf@xmath0 a jt transition temperature of 1350 k has been estimated.@xcite the paramagnetic ( pm ) susceptibility has been described by a bonner - fisher law with an exchange constant j = 190 k,@xcite indicating that the compound is a good realization of a one - dimensional ( 1d ) spin chain in the pm regime .
inelastic neutron scattering studies did reveal a spinon - excitation continuum , a clearly 1d quantum phenomenon , existing also below the nel temperature.@xcite from a structural point of view the reported relatively high tetragonal symmetry @xcite ( @xmath11 @xmath124@xmath13 ) makes kcuf@xmath0 one of the simplest systems to study .
however , the established symmetry assignment has been questioned by an x - ray diffraction investigation @xcite which suggested the existence of orthorhombic distortions in kcuf@xmath0 at room temperature with @xmath14 symmetry .
a low - temperature raman scattering study @xcite revealed a difference of spectra measured in @xmath15 and @xmath16 polarization and anomalously broad linewidths of the stretching modes , which was interpreted as evidence of a symmetry lower than @xmath11 also below the nel temperature .
although orthorhombic distortions were involved for explaining the electron spin resonance ( esr ) properties of kcuf@xmath0,@xcite discrepancies remain for the analysis of recent nqr,@xcite afm resonance,@xcite and further experimental and theoretical findings.@xcite besides , in x - ray resonant scattering@xcite of the orbital ordering ( oo ) in kcuf@xmath0 indications for a coupling of lattice and magnetic degrees of freedom above @xmath8 were found .
only recently , the esr properties for @xmath17 could be successfully explained within the tetragonal symmetry by assuming a dynamical dzyaloshinsky - moriya ( dm ) interaction related to strong oscillations of the bridging f@xmath18 ions perpendicular to the crystallographic @xmath19 axis.@xcite it was argued that these dynamic distortions freeze in at a temperature @xmath5 = 50 k , leading to an effectively lower symmetry and the occurrence of exciton - magnon sidebands in optical absorption experiments.@xcite here we report on a detailed study of the temperature dependence of the raman - active phonons in a kcuf@xmath0 single crystal tracking the symmetry reduction during the anticipated freezing of the dynamic distortion at @xmath5 = 50 k and the nel ordering at @xmath8 = 39 k. we find a large softening of the lowest lying @xmath1 mode and the @xmath2 mode by 25% and 13% between room temperature and @xmath5 , respectively .
the linewidth and the integrated intensity of these modes also exhibit anomalies at @xmath5 and @xmath20 moreover , the @xmath1 mode at about 260 @xmath6 clearly splits below @xmath5 evidencing the existence of an antiferrodistortive lattice instability in kcuf@xmath0 which leads to a symmetry reduction at @xmath5 = 50 k prior to magnetic ordering .
the single crystal was oriented by laue diffraction , cut along the ( 110 ) pseudocubic plane and mechanically polished to optical quality
. details on crystal growth are described in ref . .
the sample has previously been investigated by esr and optical spectroscopy.@xcite the raman spectra were obtained with two different experimental setups and in two geometries of experiment : ( i ) a dilor xy triple spectrometer with a liquid - nitrogen - cooled ccd detector ( quasi - backscattering geometry ) and ( ii ) a u1000 high resolution double spectrometer with rca 31034a photomultiplier ( right - angle scattering geometry ) .
the 647 nm ar / kr ion ( 5 mw output power ) and the 632.8 nm he - ne ( 25 mw output power ) lasers were used for excitation in these two setups , respectively .
temperature dependencies were obtained in variable temperature gas - flow cryostats .
polarized raman spectra of single crystal kcuf@xmath0 taken at 290 k in different scattering configurations.,width=302 ] in fig .
[ fig : fig1 ] the polarized raman spectra of single crystalline kcuf@xmath0 taken in @xmath21 , @xmath22 , @xmath23 , @xmath15 , and @xmath16 scattering configurations are shown for @xmath24 = 290 k. the number of lines and the selection rules are fully consistent with the theoretically expected raman - active normal modes@xcite of kcuf@xmath0 with tetragonal @xmath11 @xmath25 hence , the three lines in both the @xmath15 and @xmath16 spectra correspond to the three @xmath1 modes . the line observed with different intensities in @xmath21 and @xmath22 spectra
is identified as the @xmath26 mode .
the intense line observed only in the @xmath21 spectrum can be assigned to the @xmath2 mode . finally , the two lines in the @xmath23 spectra are the two @xmath27 modes . at room temperature
all lines have a lorentzian lineshape .
figure [ fig : fig2 ] shows schematically the vibrational patterns for the seven raman - active modes of each symmetry ( @xmath26 , @xmath2 , @xmath27 , and @xmath1 ) of kcuf@xmath0 derived from the @xmath11 space group . the observed spectra and mode assignments are in agreement with previously reported data at 10 k.@xcite a direct comparison of our data at 4 k and 290 k with ref . and theoretical estimates@xcite is presented in tab .
[ fig : tab1 ] . in general
, there is a good agreement between the corresponding values except for the @xmath27(1 ) mode with a frequency of 240.4 @xmath6 observed in our experiments in contrast to a somewhat higher frequency of 265.8 @xmath6 in ref . .
the second discrepancy is that the lines assigned to @xmath1(1,2 ) and @xmath2 are almost two times broader in the low - temperature raman spectra of ref . .
the phonon lines of @xmath26 and @xmath27 symmetry have large linewidths in comparison with the other modes . in fig .
[ fig : a1goverview ] we show the temperature dependent parameters for the @xmath26 mode as an example .
the @xmath26 and @xmath27 modes , aside from their broadened lineshape , show no anomalous behavior . in the full temperature range
they exhibit a hardening of 1 - 2@xmath4 .
moreover , we observe quasielastic scattering in @xmath22 configuration , which is a general feature in low - dimensional spin systems@xcite and should only be observed in intra - chain scattering configuration , i.e. with the light polarization parallel to the effective chain direction .
this quasielastic scattering in kcuf@xmath0 has been investigated in detail by yamada and onda previously@xcite and will not be further considered in our work .
raman allowed phonon modes for the kcuf@xmath0 with @xmath11 structure .
the degeneracy of the @xmath1 modes is indicated using solid and dotted arrows.,width=302 ] c@*7c mode & & + & 290k & 4 k & 10 k ( ref . ) & calculated ( ref . ) & 290k & 4k & 10 k ( ref . )
+ @xmath26 & 367.3 & 373.5 & 374.8 & 398 & 23.1 & 4.9 & 9.2 + @xmath2 & 81.6 & 70.9 & 72.8 & 100 & 7.1 & 0.9 & 1.6 + @xmath27(1 ) & 240.4 & 245.2 & 265.8 & 259 & 30.8 & 8.7 & 7.0 + @xmath27(2 ) & 554.8 & 561.3 & 563.0 & 586 & 22.6 & 9.1 & 9.1 + @xmath28(1 ) & 63.0 & 47.4 & 53.2 & 50 & 5.8 & 0.7 & 3.0 + @xmath28(2 ) & 262.9 & 260.8 & 261.6 & 136 & 7.5 & 1.7 & 3.0 + @xmath28(3 ) & 132.3 & 129.3 & 131.2 & 268 & 7.5 & 1.6 & 1.6 + parameters of the @xmath26 mode : temperature dependence of ( a ) the squared eigenfrequency @xmath29 , ( b ) the fwhm linewidth , and ( c ) the bose corrected integrated intensity .
lines are to guide the eye.,width=302 ] in the following we will focus on the temperature dependence of the modes @xmath30 , @xmath31 and @xmath2 .
the @xmath1(1 ) mode reportedly exhibits a weak splitting at 10 k only when measured in @xmath16-configuration .
in contrast , the @xmath1(2 ) mode shows a splitting only when measuring in @xmath15-configuration .
the @xmath1(3 ) mode , which corresponds to a vibration of k@xmath32 ions , shows no splitting in either of the two configurations.@xcite the @xmath1(1 ) and @xmath1(2 ) modes correspond to shearing vibrations of the f@xmath18 ions which involves a displacement of the fluorine ions away from the cu - f - cu bonding lines , while the @xmath2 mode corresponds to a tilting motion of the f@xmath18 ions around the central cu atom ( see fig .
[ fig : fig2 ] ) .
as such displacements are thought to be the origin of the dynamical dm interaction which allows to understand the esr and antiferromagnetic resonance properties , we expect that these modes are strongly related to the proposed freezing of the dynamic fluorine displacements below @xmath5 = 50 k.@xcite temperature dependent raman spectra of the @xmath1(1 ) and @xmath1(2 ) modes taken in the @xmath16 scattering geometry.,width=302 ] parameters of the @xmath33 mode in the @xmath16 scattering geometry : temperature dependence of ( a ) the squared eigenfrequency @xmath34 together with a fit using eq .
[ eq : softmode ] , ( b ) the fwhm linewidth with a fit using eq .
[ eq : linewidth ] dashed and dotted lines : first and second terms in eq .
[ eq : linewidth ] , respectively , and ( c ) the bose corrected integrated intensity ( solid line is a guide to the eye ) .
the insets highlight the data for @xmath3560 k.,width=302 ] parameters of the @xmath36 mode in the @xmath16 scattering geometry : temperature dependence of ( a ) the squared eigenfrequency @xmath34 together with a fit using eq .
[ eq : softmode ] , ( b ) the fwhm linewidth with a fit using eq .
[ eq : linewidth ] dashed and dotted lines : first and second terms in eq .
[ eq : linewidth ] , respectively , and ( c ) the bose corrected integrated intensity ( line is a guide to the eyes ) .
the insets highlight the data for @xmath3560 k.,width=302 ] parameters of the @xmath2 mode : temperature dependence of ( a ) the squared eigenfrequency @xmath34 together with a fit using eq .
[ eq : softmode ] , ( b ) the fwhm linewidth with a fit using eq .
[ eq : linewidth ] dashed and dotted lines : first and second terms in eq .
[ eq : linewidth ] , respectively , , and ( c ) the bose corrected integrated intensity and a solid line to guide the eye.,width=302 ] indeed , when looking at the raman data of the @xmath1(1 ) and @xmath1(2 ) vibrational modes in @xmath16 configuration shown in fig .
[ fig : fig3 ] , an anomalous softening of both modes is observed for @xmath37 50 k. while the frequency shift of the @xmath1(2 ) is only about 2 @xmath6 , the low - energy @xmath1(1 ) mode exhibits a frequency shift of about 16 @xmath6 .
this corresponds to a softening of 1% and 25% with respect to the room temperature eigenfrequency . plotting the square of the eigenfrequency @xmath38 of the @xmath1(1 )
phonon mode as a function of temperature in fig .
[ fig : aeg1yzoverview](a ) reveals a linear behavior for @xmath37 50 k , which can be understood in terms of a soft - mode behavior indicative of a structural phase transition expected at @xmath39 where one expects@xcite @xmath40 the fit shown in fig .
[ fig : aeg1yzoverview](a ) yields @xmath41 = 6.8 @xmath42/k and a virtual transition temperature of @xmath39 = -291 k. although the negative sign indicates that the occurrence of the structural phase transition is very unlikely , the energy scale of this virtual transition temperature is close to the orbital - ordering transition temperature @xmath43 350 k calculated by assuming a purely electronic superexchange mechanism.@xcite we believe that the softening of the @xmath1(1 ) phonon mode is due to the dynamic nature in the displacement of the apical fluorine ions away from the @xmath19 axis , which manifests itself in an anomalously large thermal displacements parameter@xcite and the occurrence of a dynamical dzyaloshinsky - moriya ( dm ) interaction.@xcite as a prerequisite for the latter , the characteristic time of the dynamic distortions must be large compared to the time scale of the exchange interaction and the amplitude of these distortions must be high.@xcite this is the case for low - lying optical modes contributing to the oscillation of the f@xmath18 ions with the tendency to soften to low temperatures , exactly like the @xmath1(1 ) mode . in this scenario
the displacement of the fluorine ions freezes with decreasing temperature and becomes static at @xmath5= 50 k.@xcite below 50 k we observe a deviation from this softening behavior and the frequency levels off in the magnetically ordered state . the temperature dependence of the phonon line widths full width at half maximum ( fwhm ) for @xmath1(1 )
is shown in fig .
[ fig : aeg1yzoverview](b ) . above 50
k the linewidth data can be described ( solid line ) using : @xmath44 where @xmath45 is the contribution arising from phonon anharmonic interactions in crystalline solids , with zone center modes decaying into pairs of phonons with equal and opposite wave vectors .
@xmath46 is the contribution to the total linewidth arising from the dynamic deviation of the f@xmath18 ions away from the @xmath19 axis , @xmath47 is a potential barrier , and @xmath48 and @xmath49 are constants . the data can be described very well over the temperature range 50 290 k by eq .
[ eq : linewidth ] yielding a energy @xmath47 = 56 k very close to the temperature @xmath5 = 50 k where the dynamic displacements are proposed to become static.@xcite below @xmath5 = 50 k the width of the phonon line decrease nearly linearly with temperature .
the ( bose corrected ) integrated intensity of the @xmath1(1 ) mode shown in fig .
[ fig : aeg1yzoverview](c ) increases with decreasing temperature and reaches a maximum at @xmath5 ( see inset of fig . [ fig : aeg1yzoverview](c ) ) and a minimum just below @xmath8 reflecting distinct changes of the polarizability of this mode at these temperatures .
the corresponding parameters for the @xmath1(2 ) and the @xmath2 mode are plotted in fig .
[ fig : eg2yzoverview ] and fig .
[ fig : b1goverview ] , respectively .
similar to the @xmath1(1 ) mode these modes exhibit a soft mode behavior with @xmath41=3.99 @xmath42/k and @xmath39=-17016 k for the @xmath1(2 ) and @xmath41=6.57 @xmath42/k and @xmath39=-722 k for the @xmath50(1 ) mode .
while for the @xmath50(1 ) with a softening of about 13% with respect to room temperature the virtual transition temperature is still reasonable , the value for the @xmath1(2 ) mode appears not to be of physically meaningful due to the moderate softening of only 1 - 2% .
note that in other fluorides with rutile structure like mnf@xmath51 , nif@xmath51 , and fef@xmath51 virtual transition temperatures of -1240 k , -1700 k , and -1780 k have been derived from the softening of raman modes , respectively.@xcite the linewidth of both modes can again be described using eq.[eq : linewidth ] and @xmath47 = 49 k. the intensities of both modes start to decrease below 50 k , but for the @xmath1(2 ) mode the intensity levels off and becomes almost constant below @xmath8 .
although clear anomalies of these modes associated with @xmath5 and @xmath8 have been observed , we could not observe the splitting of the @xmath1(1 ) mode in @xmath16-configuration reported in ref . at 10 k.
hence , we tried to reproduce the reported splitting of the @xmath1(2 ) mode in @xmath15-configuration and trace its temperature dependence . the obtained spectrum at 3.5 k is shown in fig .
[ fig : eg2_xz_spectrum ] and a weak additional mode at the high - frequency side of the @xmath1(2 ) is clearly visible compared to the data at t=50 k in the same figure .
the solid line corresponds to a fit with two lorentzian lineshapes ( thin and dashed , latter shifted ) . as shown in fig .
[ fig : eg2_xz_freq2 ] , the appearance of this additional mode coincides with @xmath5 suggesting a splitting of the @xmath1(2 ) mode in agreement with the scenario of a symmetry reduction at @xmath5 suggested previously.@xcite nevertheless , we have to point out that this splitting of about 6 @xmath6 is larger than the reported one of about 1 @xmath6 and appears on the high - energy flank in contrast to the one reported by ueda and coworkers,@xcite which appears on the low - energy side of the original @xmath1(2 ) mode .
these discrepancies can not be easily explained and may be due to the different samples used for our work .
we would like to point out that in a recent raman study a splitting of the @xmath1(2 ) similar to our data has been reported.@xcite raman spectrum of the @xmath1(2 ) in @xmath15 scattering geometry at 3.5 k ( crosses ) .
the solid line and the dashed line ( shifted ) are the result of fitting the data with two lorentzian lines . for reference the data at 50 k ( open circles ) is shown with a single lorentzian line . ,
width=302 ] a static displacement of fluorine ions away from the @xmath19 axis at temperatures @xmath52 assumes the lowering of the kcuf@xmath0 crystal symmetry .
if the symmetry is lower than @xmath53 , a removal of the @xmath1 modes degeneracy and the appearance of extra lines in the raman spectra is expected .
the observed splitting of the @xmath1(2 ) mode in @xmath15 configuration confirms this scenario , alone , this information is not sufficient to determine the low - temperature symmetry .
additional evidence has been obtained by x - ray scattering where a splitting of a bragg reflection associated with gdfeo@xmath0 type distortions has been found below 50 k.@xcite lee and coworkers also suggested that the observed softening of @xmath1 and the @xmath2 modes is related to the finite spin correlation lengths which are inherent to low - dimensional magnets.@xcite such effects of spin - phonon coupling are well established and occur , e.g. , in frustrated magnetic systems without orbital degrees of freedom.@xcite temperature dependence of the eigenfrequency of the @xmath1(2 ) and the split modes below @xmath5 in @xmath15 scattering geometry .
lines are drawn to guide the eye.,width=302 ] in this respect it is worth highlighting that in kcuf@xmath0 the spin - spin relaxation time as measured by the esr linewidth can only be explained by assuming dynamic lattice distortion of the type associated with the anomalous raman modes.@xcite it was also reported in ref . that the temperature dependence of the esr linewidth @xmath54 can be described by @xmath55 with an activation energy @xmath56=114 k which corresponds approximately to @xmath57 , two times the potential barrier derived from the temperature dependence of the linewidths of the anomalous raman modes .
this intricate feedback between spin , lattice , and possibly the orbital degrees of freedom has to be disentangled and the following questions arise and still need to be clarified : ( i ) is the softening of the raman modes directly related to the spin - spin correlations of the quasi - one dimensional spin chain kcuf@xmath0 ?
( ii ) is there a relation between the esr spin - spin relaxation time dominated by the dynamical dzyaloshinsky - moriya interaction and the linewidth of the anomalous raman modes ?
( iii ) how and on which time scale do the orbital degrees of freedom couple to the lattice and spin fluctuations in the system ? we hope that our study will stimulate further theoretical efforts in this direction .
to sum up , temperature - dependent raman spectra of single crystalline kcuf@xmath0 show a strong softening of the lowest - lying @xmath1(1 ) and the @xmath2 mode for @xmath7 . both of these modes and the @xmath1(2 ) mode ( at about 260 @xmath6 ) exhibit anomalies at the characteristic temperature @xmath5=50 k. in @xmath15 scattering configuration the @xmath1(2 )
doublet clearly splits with a splitting of about 6 @xmath6 .
the temperature dependence of the linewidth of these modes yields an activated behavior with an energy @xmath58 50 k corresponding to @xmath5 .
we ascribe this anomalous behavior and the observed splitting to an antiferrodistortive lattice instability due to strong dynamic displacements of the f@xmath18 ions away from the cu - f - cu bonding line along the @xmath19 axis .
these displacements are strongly influencing the spin - spin relaxation by allowing for a dynamical dzyaloshinsky - moriya interaction .
they become static for t@xmath59@xmath5 .
we like to thank m. v. eremin and b. lake for useful discussions . v.g . and o.a .
acknowledge the russian - ukrainian grant 2009 - 9 for partial support
acknowledges support by b - igsm .
we also acknowledge support by the dfg via trr80 and via le 967/6 - 1 and the swiss nsf through nccr manep . | in the orbitally ordered , quasi - one dimensional heisenberg antiferromagnet kcuf@xmath0 the low - energy @xmath1 and @xmath2 phonon modes show an anomalous softening ( @xmath325@xmath4 and @xmath313@xmath4 ) between room temperature and the characteristic temperature @xmath5 = 50 k. in this temperature range a freezing - in of f ion dynamic displacements is proposed to occur .
in addition , the @xmath1 mode at about 260 @xmath6 clearly splits below @xmath5 .
the width of the phonon lines above @xmath5 follows an activated behavior with an activation energy of about 50 k. our observations clearly evidence a reduction of the structural symmetry below @xmath5 and indicate a strong coupling of lattice and spin fluctuations for @xmath7 . |
solid hydrogen h@xmath0 is expected to become metallic at compressions higher than @xmath3 gpa [ ] . in fact , experimental signatures of the long - sought insulator - to - metal phase transition remain elusive up to pressures of @xmath4 gpa [ ] .
accepted pressure - induced mechanisms by which the metallicity of hydrogen can be enhanced involve atomization of h@xmath0 molecules and partial filling of electronic @xmath5 molecular levels due to charge transfer from or band hybridization with other chemical species [ ] .
already in the earlier 90 s loubeyre _
et al . _ , based on the disappearance of the raman ( r ) vibron mode and the darkening of the material , claimed to observe metallization of the ar(h@xmath0)@xmath0 compound when compressed in the diamond - anvil - cell ( dac ) up to @xmath6 gpa [ ] .
the stable room - temperature ( rt ) phase structure of this compound was identified with the hexagonal c14 laves structure typified by the mgzn@xmath0 crystal ( space group : @xmath7 ) .
strikingly , posterior synchrotron infrared ( ir ) measurements did not show evidence of molecular bonding instabilities nor metallic drude - like behavior up to at least @xmath8 gpa [ ] .
subsequently , bernard _ et al .
_ suggested that activation of h@xmath0 dissociation processes and corresponding development of metallization in ar(h@xmath0)@xmath0 could occur via a solid - solid phase transition of the mgzn@xmath0 @xmath1 alb@xmath0 ( space group : @xmath7 ) type at pressures already within the reach of dac capabilities [ ] .
however , recent _ ab initio _ work done by matsumoto _
demonstrates that the onset of metallicity in the alb@xmath0 structure commences at pressures significantly higher than in pure bulk h@xmath0 [ ] . in view of the growing interest on hydrogen - rich van der waals ( vdw ) compounds under pressure [ ] , partly motivated by the hydrogen - storage problem , and of the unresolved discrepancies described above ,
we have conducted a theoretical study on ar(h@xmath0)@xmath0 under extreme @xmath9 conditions using first - principles density functional theory ( dft ) calculations and _ ab initio _ molecular dynamics simulations ( aimd ) . in this letter , we present results showing that at low temperatures and pressures up to @xmath10 gpa the ar(h@xmath0)@xmath0 crystal stabilizes in the cubic c15 laves structure typified by the mgcu@xmath0 solid ( space group : @xmath11 ) .
this structure has not been considered in previous works [ ] though its probable relevance to ar(h@xmath0)@xmath0 was pointed out recently [ ] . on the light of first - principles enthalpy and phonon calculations , we propose a temperature - induced ( pressure - induced ) phase transition of the mgcu@xmath0 @xmath1 mgzn@xmath0 ( mgzn@xmath0 @xmath1 mgcu@xmath0 ) type that may clarify the origin of the discrepancies between the sets of r and ir data .
furthermore , in the high-@xmath12 regime ( @xmath13 gpa ) we find that a metallic hydrogen - rich liquid can be stabilized at temperatures of @xmath14 k wherein h - h coordination features render molecular dissociation activity . by means of aimd simulations , we estimated an upper bound of the melting curve @xmath15 of ar(h@xmath0)@xmath0 and found a negative @xmath16 slope spanning over the interval @xmath2 gpa .
our simulations show that the lattice composed by h@xmath0 molecules melts at temperatures significantly lower than the lattice of ar atoms does , so leading to stable mixtures of coexisting liquid h@xmath0 and solid ar over wide @xmath9 ranges .
we propose an argument based on this atypical physical behavior to explain the cause of the estimated negative @xmath16 slope .
phase diagram of ar(h@xmath0)@xmath0 under pressure .
mgzn@xmath0 - mgcu@xmath0 and mgcu@xmath0 - alb@xmath0 phase boundaries are sketched according to the results and arguments presented in the text .
thermodynamic states at which x - ray , r and ir vibron measurements were carried out are indicated : [ ] x - ray = @xmath17 , [ ] r = @xmath18 , [ ] ir = @xmath19 and [ ] ir = @xmath20 .
_ inset _ : enthalpy difference per particle of the mgcu@xmath0 ( solid line ) and alb@xmath0 ( dashed line ) structures with respect to the mgzn@xmath0 laves phase as function of pressure at zero temperature . ]
our calculations were performed using the all - electron projector augmented wave method and generalized gradient approximation of wang and perdew as implemented in the vasp code [ ] .
dense monkhorst - pack special @xmath21-point meshes [ ] for sampling of the first brillouin zone ( ibz ) and a cutoff energy of @xmath3 ev were employed to guarantee convergence of the total energy per particle to within @xmath22 mev . in particular , we used @xmath23 , @xmath24 and @xmath25 @xmath21-point grids for calculations on the perfect unit cell corresponding to the mgcu@xmath0 , mgzn@xmath0 and alb@xmath0 crystal structures , respectively .
all the considered crystal structures were relaxed using a conjugate - gradient algorithm and imposing the forces on the particles to be less than @xmath26 ev / .
the phonon frequencies in our calculations were obtained using the small - displacement method [ ] over the unit cells ( @xmath27-point phonon frequencies ) and large supercells containing @xmath28 atoms . _
ab initio _ molecular dynamics simulations were carried out in the canonical ensemble @xmath29 using bulk supercells of ar(h@xmath0)@xmath0 containing @xmath30 atoms ( @xmath27-point sampling ) . at
given pressure , the dynamical properties of the system were sampled at @xmath3 k intervals from zero - temperature up to the melting curve of pure ar .
temperatures were maintained using nos - hoover thermostats .
a typical aimd simulation consisted of @xmath31 ps of thermalization followed by @xmath32 ps over which statistical averages were taken .
it is worth noticing that we recently used a very similar computational approach to the one described here to study the energetic and structural properties of ar(he)@xmath0 and ne(he)@xmath0 crystals under pressure [ ] , and that very recent experiments [ ] have confirmed the validity of our predicted @xmath33 curves . )
@xmath0 in the mgcu@xmath0 ( solid line ) and mgzn@xmath0 ( dashed line ) crystal laves structures .
experimental data is from work [ ] ._inset _ : calculated pressure - dependence of the h - h intermolecular distance in the mgcu@xmath0 ( solid line , solid dots ) and mgzn@xmath0 ( dashed line , solid triangles ) crystal structures .
volumes and distances are in units of @xmath34/particle and , respectively . ]
a series of candidate structures were considered in our _ ab initio _ enthalpy @xmath36 calculations ( rutile , fluorite , mgni@xmath0 , etc . ) however only the mgzn@xmath0 , mgcu@xmath0 and alb@xmath0 structures turned out to be energetically competitive so the following analysis concentrates on these ones .
ignoring quantum zero - point motion effects , we found that ar(h@xmath0)@xmath0 is energetically more stable in the mgcu@xmath0 structure than in either the mgzn@xmath0 or alb@xmath0 structures at pressures @xmath37 gpa ( see figure [ fig : pd - h ] ) . also we predicted that at @xmath38 gpa the solid transitates from the mgcu@xmath0 to the alb@xmath0 structure .
the calculated zero - temperature equation of state of ar(h@xmath0)@xmath0 in the mgcu@xmath0 structure displays very good agreement with respect to the available experimental data [ ] ( see figure [ fig : eos ] ) .
for instance , at @xmath39 and @xmath40 @xmath34/particle we obtain a pressure of @xmath41 and @xmath42 gpa , respectively , to be compared with the experimental values @xmath43 and @xmath44 gpa . it must be noted that the estimated equation of state and pressure - dependence of the h - h intermolecular bond distance of the mgzn@xmath0 and mgcu@xmath0 phases appear to be indistinguishable within the numerical uncertainty in our calculations ( namely , @xmath45 gpa and @xmath46 - see figure [ fig : eos]- ) .
since hydrogen is a very light molecule , quantum zero - point motion corrections must be included in the calculations .
customarily this is achieved using quasi - harmonic approaches that involve estimation of the vibrational phonon frequencies . in following this procedure
, we found that ar(h@xmath0)@xmath0 in the mgzn@xmath0 structure always exhibits imaginary @xmath27-phonon frequencies at pressures below @xmath47 gpa .
this means that the mgzn@xmath0 structure is mechanically unstable , at least , at low temperatures . on the other hand
, we found that the mgcu@xmath0 structure is perfectly stable at all the studied pressures .
our numerical tests showed that this result does not depend on the approximation of the exchange - correlation functional used [ ] . in figure
[ fig : phonons ] , we show the phonon frequency spectra of ar(h@xmath0)@xmath0 in the mgcu@xmath0 crystal structure calculated along a @xmath21-point path contained within the ibz and at pressure @xmath6 gpa . )
@xmath0 in the mgcu@xmath0 crystal structure at @xmath48 @xmath34/particle and @xmath49 gpa ( corresponding molecular vibron frequencies are not shown ) .
the calculation was done on a supercell containing @xmath50 ar and @xmath51 h atoms and using a @xmath52 @xmath21-point grid for ibz sampling . ]
since the x - ray resolved rt crystal structure of ar(h@xmath0)@xmath0 is mgzn@xmath0 [ ] , we carried out a series of aimd simulations of this and the mgcu@xmath0 structure at @xmath53 k in order to explore their stability .
in fact , analysis of our simulations based on estimation of the averaged mean - squared displacement and position - correlation function [ ] showed that both mgcu@xmath0 and mgzn@xmath0 structures are mechanically stable at rt . on view of these results and the reported x - ray data
, we concluded that a temperature - induced mgcu@xmath0 @xmath1 mgzn@xmath0 transition occurs at fixed @xmath12 ( or equivalently , a pressure - induced mgzn@xmath0 @xmath1 mgcu@xmath0 transition at fixed @xmath35 - see figure [ fig : pd - h ] - ) .
this temperature - induced transformation can be understood in terms of entropy : ar(h@xmath0)@xmath0 in the mgzn@xmath0 structure is highly anharmonic so ionic entropy contributions to the total free - energy stabilize this structure over the mgcu@xmath0 phase with raising temperature .
in fact , ir measurements displayed an splitting of the vibron mode in the interval @xmath54 k [ ] . this sppliting can be induced by the degenerancy removal of the infrarred @xmath55 mode due to the decrease of the crystal quality caused by precursor effects of the mgzn@xmath0 @xmath1 mgcu@xmath0 transition . )
@xmath0 in the mgzn@xmath0 , mgcu@xmath0 and alb@xmath0 structures as function of pressure .
experimental data from [ ] ( @xmath17 ) and [ ] ( @xmath18 ) are shown . ]
the causes of the aforementioned r - ir experimental disagreements can be rationalized in the light of the foreseen mgcu@xmath0 @xmath1 mgzn@xmath0 transition , as we explain in what follows . in figure
[ fig : vibron ] , we enclose the calculated r and ir vibron frequencies of ar(h@xmath0)@xmath0 in the mgzn@xmath0 , mgcu@xmath0 and alb@xmath0 structures as function of pressure . at high pressures , it is shown that the @xmath56 vibron line corresponding to the mgzn@xmath0 and mgcu@xmath0 structures are practically identical , whereas they settle appreciably below the line estimated for the alb@xmath0 structure , and follow closely the low-@xmath35 r results [ ] . these outcomes together with the energy and phonon results already presented , led us to think that the sudden dissapearance of the r vibron mode observed at high-@xmath12 might be related to the mgcu@xmath0 @xmath1 mgzn@xmath0 transition unravelled in this letter and not to pressure - induced metallic - like behavior [ ] or mgzn@xmath0 @xmath1 alb@xmath0 phase transition [ ] as previously suggested .
in fact , accurate electronic density of states ( dos ) analysis performed on the perfect mgzn@xmath0 , mgcu@xmath0 and alb@xmath0 crystal structures show no closure of the electronic band gap up to compressions of at least @xmath57 gpa .
furthermore , the @xmath58 vibron line that we calculated for the mgzn@xmath0 structure ( see figure [ fig : vibron ] ) agrees closely with datchi s ir data obtained at high pressures .
this accordance is coherent if one considers that the series of ir experiments were realized near the edge of the mgcu@xmath0-mgzn@xmath0 phase boundary , as we sketch in figure [ fig : pd - h ] .
conclusive experiments confirming the validity of our statements might consist of new series of x - ray and ir measurements performed at temperatures below @xmath59 k and high pressures ; according to our predictions , the vibron signature will eventually disappear at the crossing with the mgcu@xmath0-mgzn@xmath0 phase boundary since ir vibron modes in the mgcu@xmath0 structure are inactive . ) @xmath0 under pressure ( solid line ) .
melting states obtained in the aimd simulations are indicated with @xmath18 and corresponding error bars amount to @xmath60 k. thermodynamic states at which the lattice of hydrogen molecules is observed to melt in the simulations are represented by @xmath20 ( fitted to a dashed line ) .
the melting line of pure ar [ ] ( long - dashed and dotted line ) and h@xmath0 [ ] ( dashed and dotted line ) are shown for comparison . ]
an intriguing physical phenomenon has been recently predicted and subsequently observed in h@xmath0 under pressure .
it consists in the appearance of a maximum peak on its melting line followed by a negative @xmath16 slope [ ] .
this phenomenon have been explained in terms of subtle changes on the intermolecular forces due to compression instead of more familiar arguments like promotion of valence electrons to higher energy orbitals or occurrence of molecular dissociation processes .
motivated by these interesting findings on hydrogen , we investigated the dynamical properties of ar(h@xmath0)@xmath0 at high-@xmath12 and high-@xmath35 in the search of similar physical manifestations and to provide further understanding of vdw compounds in general . there
exist several well - established techniques by which one can determine the melting line of a material ; these essentially base on solid - liquid phase coexistence simulations and/or gibbs free - energy calculations [ ] .
application of these methods at quantum first - principles level of description , however , turns out to be computationally very intensive and laborious .
a simulation approach that has proved successful in reproducing general melting trends in materials at affordable computational cost , including experimental negative melting slopes , is the ` heat - until - it - melts ' method [ ] . this technique allows for estimation of a precise upper bound of the solid - liquid phase boundary of interest .
it must be stressed that we did not pursue accurate calculation of the melting line of ar(h@xmath0)@xmath0 but to identify possible anomalous effects on it . )
@xmath0 calculated at @xmath13 gpa ( @xmath48 @xmath34/particle ) and different temperatures .
_ inset _ : averaged radial h - h pair distribution function g@xmath61 obtained for the liquid ( solid line ) and solid ( dashed line ) phases ( distance is in units of ) . snapshots of liquid ( l ) and solid ( s ) configurations as generated in the aimd simulations ; dissociated h@xmath0 molecules are observed in l. ] results from our aimd simulations are shown in figure [ fig : ub - melting ] and can be summarized as follows : ( i ) the lattice of h@xmath0 molecules melts at temperatures significantly lower than the whole crystal does so leading to mixtures of liquid h@xmath0 and solid ar over wide @xmath9 ranges ; ( ii ) the fusion of the h@xmath0 lattice occurs at temperatures very close to the melting line of pure hydrogen while ar(h@xmath0)@xmath0 as a whole practically reproduces the melting behavior of pure ar up to @xmath62 gpa ; ( iii ) the value of the estimated @xmath16 slope is negative within the pressure interval @xmath2 gpa whereas positive elsewhere .
results ( i ) and ( ii ) can be interpreted in terms of similar arguments than recently disclosed in solid mixtures of ne - he and ar - he , namely : lattices composed of same - species particles effectively behave like not interacting one with another but following alike physical trends ( bond distance , compressibility , etc . ) than found in their respective pure system [ ] . in fact , we recently conjectured the superior energy stability of the mgcu@xmath0 structure over mgzn@xmath0 in ar(h@xmath0)@xmath0 , as rigorously demonstrated here , using this type of reasoning and crystal symmetry arguments [ ] .
regarding result ( iii ) , this is in itself a manifestation of a very peculiar and intriguing physical phenomenon . in general , systems presenting negative melting slope are characterized by open crystalline structure ( water and graphite ) , surpassing promotion of valence electrons in the fluid phase ( alkali metals ) or continuous changing interparticle interactions ( molecular hydrogen ) . in the present case either open crystalline structure or electronic band promotion effects ( energetically prohibitive ) can be ruled out , so changes in the interparticle interactions holds as the likely cause .
the question to be answered next then is : what is the exact nature of these changing interactions ? , or more precisely , do they relate to dramatic changes in electronic structure ?
are the ionic degrees of freedom and effective decoupling of the ar and h@xmath0 lattices crucial to them ? in order to detect possible electronic phase transitions and/or molecular dissociation processes , we performed meticulous averaged dos and h - h radial correlation function ( @xmath63 ) analysis over the atomic configurations generated in the aimd runs . within the thermodynamic range @xmath2 gpa and @xmath64 k ( where @xmath65 ) , we found that ar(h@xmath0)@xmath0 is always an insulator material , either solid or liquid , where h@xmath0 molecules remain stable
therefore , dramatic changes in electronic structure can be discarded as the originating mechanism behind the alternating sign of @xmath16 . actually , on the contrary case , one would have expected the value of @xmath16 to be negative also beyond @xmath66 gpa since in principle there is not reason to think that the originating electronic effects go missing at higher-@xmath12 .
interestingly , at compressions above @xmath67 gpa and temperatures high enough for the system to melt we did observe closure of the electronic energy band gap ; according to @xmath63 analysis and visual recreation of the aimd configurations , this electronic phase transition is a consequence of emergent h@xmath0 dissociation processes . in figure
[ fig : dos - hh ] , we enclose dos and @xmath63 results that illustrate this insulator - to - metal phase transition .
it is worth noticing that this result is consistent with shock - wave compression experiments [ ] and _ ab initio
_ investigations [ ] performed on pure hydrogen . with regard to the predicted negative melting slope ,
ionic effects are the likely subjacent cause .
considering results ( i ) and ( ii ) above , one can envisage a plausible explanation of this phenomenon : since the lattice of h@xmath0 molecules melts at temperatures much lower than the lattice of ar atoms does , collisions between diffusive h@xmath0 molecules and localized ar atoms act like effective thermal - like excitations on the last that ultimately provoke the global melting of the system at temperatures below those of pure ar .
h@xmath0 itself presents negative melting slope beyond @xmath62 gpa so this effect is echoed , and enhanced significantly , in the melting line of ar(h@xmath0)@xmath0 ( see figure [ fig : ub - melting ] ) . beyond @xmath66 gpa ,
the stability of the ar lattice becomes further reinforced so the effect of h@xmath0 collisions there is just to deplete the slope of the global melting curve in comparison to that of pure ar . in order to test the validity of such hypothesis , we performed a series of aimd simulations at @xmath68 gpa ( @xmath69 @xmath34/particle ) in which we kept the lattice of h@xmath0 molecules _
frozen _ and left the lattice of ar atoms to evolve . under these conditions we found that fusion of the ar structure occurs at temperatures @xmath70 k above that of pure ar , hence the original melting mechanism proposed seems to be corroborated .
analogous melting behavior than described here can be expected in other rare gas - h@xmath0 compounds , like xe(h@xmath0)@xmath71 , and maybe also in sih@xmath72(h@xmath0)@xmath0 and ch@xmath72(h@xmath0 ) .
to summarize , we have studied the behavior of ar(h@xmath0)@xmath0 under pressure at low and high temperatures using computational first - principles techniques . as results ,
we have unravelled ( i ) temperature(pressure)-induced solid - solid phase transitions that may resolve the existing discrepancies between the sets of r and ir experimental data , and ( ii ) an anomalous melting phenomenon consisting of negative melting slope .
the atypical melting line of ar(h@xmath0)@xmath0 can be understood in terms of the decoupling of h@xmath0 and ar ionic degrees of freedom and of coexistence of same - species liquid and solid phases .
metallization of liquid ar(h@xmath0)@xmath0 is predicted at high @xmath9 conditions .
30 m. stdele and r. martin , phys .
lett * 84 * , 6070 ( 2000 ) ; p. loubeyre , f. occelli and r. letoullec , nature * 416 * , 613 ( 2002 ) c. narayana , h. luo , h. orloff and a. l. ruoff , nature ( london ) * 393 * , 46 ( 1998 ) p. loubeyre , r. letoullec and j - p .
pinceaux , phys .
* 72 * , 1360 ( 1994 ) t. a. strobel , m. somayazulu and r. j. hemley , phys .
lett . * 103 * , 065701 ( 2009 ) e. zurek , r. hoffmann , n. w. ashcroft , a. r. oganov and a. o. lyakhov , pnas * 106 * , 17640 ( 2009 ) f. datchi , p. loubeyre , r. letoullec , a. f. goncharov , r. j. hemley and h. k. mao , bull .
phys . soc .
* 41 * , 564 ( 1996 ) s. bernard , p. loubeyre and g. zrah , europhys
. lett . * 37 * , 477 ( 1997 ) n. matsumoto and h. nagara , j. phys .
: condens .
matter * 19 * , 365237 ( 2007 ) m. somayazulu
_ et al . _ , nature chem . * 2 * , 50 ( 2010 ) h. chacham and b. koiller , phys .
b * 52 * , 6147 ( 1995 ) c. cazorla , d. errandonea and e. sola , phys . rev .
b * 80 * , 064105 ( 2009 ) g. kresse and j. furthmller , phys . rev .
b * 54 * , 11169 ( 1996 ) ; y. wang and j. p. perdew , phys . rev .
b * 44 * , 13298 ( 1991 ) h. j. monkhorst and j. d. pack , phys . rev .
b * 13 * , 5188 ( 1976 ) g. kresse , j. furtmller and j. hafner , europhys .
* 32 * , 729 ( 1995 ) d. alf , comp .
. comm . * 180 * , 2622 ( 2009 ) h. fukui , n. hirao , y. ohishi and a. q. r. baron , j. phys . : condens .
matter * 22 * , 095401 ( 2010 ) d. m. ceperley and b. i. alder , phys .
* 45 * , 566 ( 1980 ) l. voadlo , d. alf , m. j. gillan , i. g. wood , j. p. brodholt and g. d. price , nature * 424 * , 536 ( 2003 ) l. ulivi , r. bini , p. loubeyre , r. letoullec and h. j. jodl , phys .
b * 60 * , 6502 ( 1999 ) s. a. bonev , e. schwegler , t. ogitsu and g. galli , nature ( london ) * 431 * , 669 ( 2004 ) s. deemyad and i. silvera , phys .
lett . * 100 * , 155701 ( 2008 ) e. pechenik , i. kelson and g. makov , phys .
b * 78 * , 134109 ( 2008 ) m. j. gillan , d. alf , j. brodholt , l. voadlo and g. d. price , rep . prog
. phys . * 69 * , 2365 ( 2006 ) ; d. alf , m. j. gillan and g. d. price , nature * 401 * , 462 ( 1999 ) . i. tamblyn , j - y .
raty and s. a. bonev , phys .
* 101 * , 075703 ( 2008 ) j - y .
raty , e. schwegler and s. a. bonev , nature * 449 * , 448 ( 2007 ) s. t. weir , a. c. mitchell and w. j. nellis , phys .
lett . * 76 * , 1860 ( 1996 ) . | we study the structural stability and dynamical properties of ar(h@xmath0)@xmath0 under pressure using first - principles and _ ab initio _ molecular dynamics techniques . at low temperatures
, ar(h@xmath0)@xmath0 is found to stabilize in the cubic c15 laves structure ( mgcu@xmath0 ) and not in the hexagonal c14 laves structure ( mgzn@xmath0 ) as it has been assumed previously . based on enthalpy
energy and phonon calculations , we propose a temperature - induced mgcu@xmath0 @xmath1 mgzn@xmath0 phase transition that may rationalize the existing discrepancies between the sets of raman and infrared vibron measurements .
our aimd simulations suggest that the melting line of ar(h@xmath0)@xmath0 presents negative slope in the interval @xmath2 gpa .
we explain the origin of this intriguing physical phenomenon in terms of decoupling of the ar and h@xmath0 degrees of freedom and effective thermal - like excitations arising from coexisting liquid h@xmath0 and solid ar phases . |
astronomers have sought for two decades for dynamical evidence for the presence of massive black holes in galaxies by studying the dynamics of gas and stars in their nuclei .
rapid motions provide the main signature of a black hole .
if these are indeed observed , the main difficulty is to rule out alternative interpretations .
gas motions can be due to hydrodynamical processes ( inflow , outflow , turbulence , etc . ) in addition to gravity .
large stellar velocities in galactic nuclei can be the result of an overabundance of stars on radial orbits .
the primary observational tool to discriminate between different models is to obtain data of the highest possible spatial resolution . in recent years
much progress has been made in all areas of this field . for active galaxies ,
the existence of a central dark mass in at least some galaxies is now well established .
the rotation velocities of nuclear gas disks detected with the hst can be used to study the central mass distribution .
this has yielded evidence for a central dark mass of @xmath9 in m87 ( ford 1994 ; harms 1994 ) , and of @xmath10 in ngc 4261 ( ferrarese , ford & jaffe 1996 ) .
even higher spatial resolution vlba radio observations of water masers in the nucleus of the active galaxy ngc 4258 ( miyoshi 1995 ) have revealed a torus in keplerian rotation around a dark mass of @xmath11 .
for the case of ngc 4258 , there are strong theoretical arguments that this mass is indeed a black hole ( maoz 1995 ) .
the density of quasars at high redshifts suggests that many currently normal galaxies had an active phase in the past ( chokshi & turner 1992 ; haehnelt & rees 1993 ) .
hence , black holes are believed to be common in quiescent galaxies as well . in these galaxies only stellar kinematics are generally available to study the nuclear mass distribution .
the evidence for a black hole in our own galaxy is now very strong , due to proper motion measurements for individual stars near sgr a@xmath12 ( eckart & genzel 1996 , 1997 ) . for a handful of other , nearby galaxies , evidence for a central dark mass
was obtained from ground - based measurements of line - of - sight velocities ( see kormendy & richstone 1995 for a review ) , but it remained difficult to rule out all alternative models .
it was always foreseen to be a main task for the hst to improve the evidence , by providing spectra of superior spatial resolution .
stellar kinematical studies with hst became possible after the refurbishment mission in 1993 .
the first results were presented by kormendy ( 1996a , b ) , for ngc 3115 and ngc 4594 .
spectra near the nucleus with @xmath13 resolution confirmed previous arguments for black holes of @xmath14 and @xmath15 , respectively .
the quiescent e3 galaxy m32 has long been one of the best - studied black hole candidate galaxies .
the presence of a central dark mass of @xmath16@xmath17 has been argued on the basis of ground - based data with continuously increasing spatial resolution .
the most recent work indicates a central dark mass of @xmath18@xmath19 ( van der marel et al .
1994b ; qian et al .
1995 ; dehnen 1995 ; bender , kormendy & dehnen 1996 ) . however , the best ground - based kinematical data still have a spatial resolution of ` only ' @xmath20 .
goodman & lee ( 1989 ) showed that this is insufficient to rule out a cluster of dark objects , as opposed to a central black hole , on the basis of theoretical arguments .
in addition , none of the previous dynamical studies has considered axisymmetric stellar dynamical models with fully general phase - space distribution functions .
hence , it has not been shown convincingly that no plausible model can be constructed that fits the data without requiring a central dark mass .
higher spatial resolution data are therefore highly desirable .
this paper is part of a series in which we present the first hst spectra and new dynamical models for m32 .
we obtained spectra of the nuclear region with the hst / fos , and we give here a detailed description of the acquisition and analysis of these data .
dynamical models are presented in van der marel ( 1997b ) and cretton ( 1997 ) .
the main results of the project are summarized and discussed in van der marel ( 1997a ) .
the instrumental resolution of the fos has a gaussian dispersion of @xmath21 , while the wavelength scale can vary at the @xmath22 level from orbit to orbit .
a study of a low mass galaxy such as m32 ( the main body of which has a velocity dispersion of @xmath23 and a rotation velocity amplitude of @xmath24 ) is therefore significantly more complicated observationally , than that of more massive galaxies ( which have higher dispersions ) .
an additional complication is that the ` sphere of influence ' of the suspected black hole in m32 is smaller than that of most other candidate galaxies .
it is thus necessary to use the smallest fos apertures , for which it is more difficult to do an accurate target acquisition .
in addition , for a proper interpretation of the results it is necessary to determine the aperture position for each observation _ post facto _ , with an accuracy of @xmath25 . to deal with these complications it proved necessary to study the instrument and analyze the data in more than the usual detail .
parts of this paper are therefore of a somewhat technical nature .
readers interested mostly in the stellar kinematical results may wish to skip directly to section [ ss : results ] .
the paper is organized as follows .
section [ s : obsetup ] summarizes the observational setup and strategy .
section [ s : pos ] discusses the aperture positions for the observations .
section [ s : datareduc ] describes some aspects of the data reduction .
section [ s : lsf ] presents calculations of the line - spread function for each observation .
section [ s : temp ] discusses the template spectrum used in modeling the m32 spectra .
section [ s : kinematics ] discusses the kinematical analysis .
section [ s : discussion ] summarizes the main observational results . in the appendices , new calibrations and modeling
are presented of the fos aperture sizes , point - spread - function ( psf ) and line - spread - functions , for which no sufficiently accurate determinations were previously available .
we observed m32 with the red side detector of the hst / fos for nine spacecraft orbits on august 22 , 1995 ( project go-5847 ) .
the costar optics corrected the spherical aberration of the hst primary mirror .
telescope tracking was done in ` fine lock ' .
the rms telescope jitter was typically @xmath26 milli - arcsec ( mas ) .
each orbit consisted of @xmath27 minutes of target visibility time , followed by @xmath28 minutes of earth occultation .
the first two orbits were used to accurately acquire the nucleus of m32 .
subsequently , spectra were taken at various positions near the nucleus .
the g570h grating was used in ` quarter - stepping ' mode , yielding spectra with 2064 pixels covering the wavelength range from 4572 to 6821 .
periods of earth occultation were used to obtain wavelength calibration spectra of the arc lamp . in the last orbit of the observations
the fos was used in a special mode to obtain an image of the central arcsec of m32 , to verify the telescope pointing .
two different square apertures were used to obtain the spectra : the ` 0.1pair ' and the ` 0.25pair ' .
these are the smallest apertures available on the fos .
although they are paired ( two square openings separated by several arcsec ) , they were used exclusively in single - aperture mode .
all data were collected through the upper aperture .
the aperture names are based on their size in arcsec _ before _ the installation of costar .
their nominal post - costar sizes are smaller : @xmath29 square and @xmath30 square , respectively .
evans ( 1995a ) presented calibration observations in which the light throughput was measured for a point source as function of position in the apertures .
models for these observations presented in appendix [ s : appa ] indicate somewhat smaller aperture sizes : @xmath2 square and @xmath31 square , respectively .
these sizes are adopted in the remainder of the paper .
the natural coordinate system of the fos is the cartesian @xmath32 system defined in the fos instrument handbook ( keyes 1995 ) .
the apertures have their sides parallel to the axes of this system .
the grating disperses the light in the @xmath33-direction .
the @xmath32 system is ` left - handed ' when projected onto the sky . in the following
we adopt a more convenient ` right - handed ' system @xmath34 , defined through @xmath35 and @xmath36 . during the observations ,
the direction of the @xmath37-axis was fixed at a position angle of @xmath38 on the sky .
this coincides with the major axis position angle of m32 ( cf .
lauer 1992 ) .
the position of the m32 center in the hst guide star coordinate system is ra @xmath39 0h 42 m 41.82s , @xmath40 .
the positional uncertainty is dominated by that in the hst guide star catalog itself , which is @xmath20 rms .
hence , some form of acquisition is required to properly position the galaxy in the small apertures . for an extended target with a sharp and well defined brightness maximum , such as m32 ,
the method of choice is a so - called ` peak - up ' acquisition , which consists of a number of ` stages ' .
each stage adopts a rectangular grid of @xmath41 points on the sky , with inter - point spacings of @xmath42 and @xmath43 , respectively .
an fos aperture is positioned at each of the grid points , and the total number of counts is measured in some exposure time . the grid point with the most counts
is adopted as new estimate of the target position .
subsequent stages in a sequence use smaller and smaller apertures , with each stage increasing the accuracy of the target positioning .
various standard choices exist for the number of stages , and for the aperture , the grid and the exposure time for each stage in the acquisition ( keyes 1995 ) .
these standard sequences work well for point - source targets .
however , they have not been particularly well tested for extended sources , especially not for acquisitions into the small 0.1-pair aperture .
therefore , as part of the preparation of the observations , a software package was developed to simulate fos peak - up acquisitions of arbitrary targets in a monte - carlo manner ( van der marel 1995 ) .
guided by simulations done with this package , a non - standard 5-stage peak - up sequence was constructed for m32 .
the parameters of this sequence are listed in table [ t : peakup ] .
the sequence was executed with the g570h grating in place , because m32 is too bright to be acquired with the fos mirror .
the accuracy of the sequence is such that in an idealized situation ( no poisson noise , no telescope jitter , etc . )
@xmath44 and @xmath45 , where @xmath46 is the difference between the adopted pointing at the end of the acquisition and the true position of the galaxy center .
the observed intensities in the acquisition stages can be analyzed _ post facto _ , to determine the extent to which the acquisition was successful .
most important are the results of the fifth and final stage , in which the 0.1-pair aperture was sequentially positioned at the points of a @xmath47 grid on the sky , with spacing of @xmath48 between the grid points .
the left panel of figure [ f : peakup5 ] shows a grey - scale representation of the observations .
the solid dot marks the position with the highest intensity , which was adopted by the telescope software as its best estimate for the position of the galaxy center . to model the observed intensities , the ` cusp model ' for the unconvolved m32 surface brightness presented by lauer ( 1992 ) was used , which is based on pre - costar hst / wfpc images .
the point - spread - function ( psf ) of the hst+fos was approximated by a sum of gaussians , as determined in appendix [ s : appa ] .
the total magnitude at each grid point was calculated using the equations in the appendix .
the offset @xmath46 was varied to optimize the fit .
the best fit is displayed in the right panel of figure [ f : peakup5 ] .
it has @xmath49 , with a formal error in either coordinate ( based on the chi - squared surface of the fit ) of approximately 1 mas .
the target acquisition was thus successful .
an fos image of the central region of m32 was obtained in the final orbit of the observations .
the telescope was commanded to position the upper ` 1.0-pair ' aperture , which has a nominal post - costar size of @xmath50 square , on the galaxy center , after which the intensity on the photocathode was scanned with the diode array of the detector ( with no grating in the light path ) .
the scan pattern was chosen to provide @xmath51 pixels . however , the spatial resolution of the image is poor , @xmath52 , corresponding to the size of one fos diode ( koratkar 1994 ; evans 1995b ) .
the resolution can be improved by deconvolution .
figure [ f : grey_image ] shows the result of lucy deconvolving the raw image with a boxcar psf with the size of one diode .
the cross marks the galaxy center , while the dot marks the aperture center , i.e. , the position where the telescope thought the galaxy center would be .
the latter is offset from the actual galaxy center by @xmath53 ( with a formal error of 3 mas in each coordinate ) .
this offset at the end of the observations is much larger than that achieved by the target acquisition .
hence , positional errors must have accumulated during the observations .
the reason for this is not well understood .
however , the most likely cause is a drift of the telescope pointing due to thermal effects related to the heating and cooling of the spacecraft and fine guidance sensors ( fgss ) during the 14.4 hours of the observations .
such thermal effects exist ( lupie , priv . comm . ) , but were not previously reported to complicate observations with the fos ( keyes , priv .
comm . ) .
nine spectra were taken .
those obtained with the 0.1-pair aperture are referred to as s1s4 , those obtained with the 0.25-pair aperture as l1l5 .
table [ t : obsetup ] summarizes the observational setup for each spectrum .
it also lists the position with respect to the galaxy center at which the telescope was instructed to place the aperture .
these _ intended _ aperture placements are illustrated in the left panel of figure [ f : aper_plot ] . in reality , the apertures were not placed exactly at their intended positions . the actual position is a sum of two vectors : ( i ) an intentional offset from the ( telescope s estimate of the ) galaxy center ( table [ t : obsetup ] ) ; and ( ii ) an unintentional offset @xmath46 of the telescope s estimate of the galaxy center from the actual galaxy center .
the intentional offsets are applied by slewing the telescope , which should be very precise ( @xmath54 mas positional error ) . to determine the actual aperture positions one thus needs to know the offset @xmath46 for each observation .
these can be determined by modeling the observed intensities of the spectra , which are listed in table [ t : applaces ] .
the spectra s1 and s2 , taken in the subsequent orbits # 3 and # 4 , were scheduled to have identical aperture positions ( cf . table [ t : obsetup ] ) .
the same holds for the observations l4 and l5 , taken in the subsequent orbits # 8 and # 9 . in both cases , however , the observed intensities in the two spectra are significantly different .
this indicates that the aperture positions must have changed ( by @xmath55 mas , cf .
figure [ f : telpoint ] below ) between the orbits # 3 and # 4 , and between the orbits # 8 and # 9 .
there was no motion of either the grating wheel or the aperture wheel of the instrument between the different observations .
the guide star reacquisition at the beginning of each orbit is normally accurate to a few mas .
hence , the inferred pointing drifts are indeed most likely due to thermal effects on the fgs .
to proceed we make a number of simplifying assumptions about the offset @xmath46 : ( i ) the offset for the first spectrum , s1 , taken in orbit # 3 , was identical to the offset determined from the final peak - up stage ( figure [ f : peakup5 ] ) , executed in orbit # 2 ; ( ii ) the offset for the last spectrum , l5 , taken in orbit # 9 , was identical to the offset determined from the fos image ( figure [ f : grey_image ] ) , taken in that same orbit ; ( iii ) the observations s2s4 , taken in orbits # 4 , # 5 and # 6 , have a common offset ; and ( iv ) the observations l1l4 , taken in orbits # 7 and # 8 , have a common offset .
these assumptions do not allow for positional errors between orbits # 4 and # 5 , between orbits # 5 and # 6 , and between orbits # 7 and # 8 .
even though errors could have occurred , they are not required to fit the observed intensities .
the assumptions do allow for an error between orbits # 6 and # 7 , when the 0.1-pair aperture was replaced by the 0.25-pair aperture .
this is because these apertures might not be exactly concentric on the aperture wheel ( evans 1995 ) , and because the motion of the aperture wheel might have a positional non - repeatability ( dahlem & koratkar 1994 ) . with these assumptions
, the offset @xmath46 can be determined for each observation by fitting to the observed intensities , using the lauer ( 1992 ) cusp model for the m32 surface brightness , and the aperture sizes and psf determined in appendix [ s : appa ] .
the results are displayed in figure [ f : telpoint ] .
the corresponding aperture placements for the individual spectra are listed in table [ t : applaces ] and are displayed in the right panel of figure [ f : aper_plot ] .
the fits to the intensities constrain only the absolute value of @xmath56 , and not its sign .
this introduces an ambiguity for l1l4 , which was resolved by adopting the same sign for @xmath56 as measured in figure [ f : grey_image ] ( so as to minimize the size of the positional error between orbits # 8 and # 9 ) . at the positions of the apertures ( along the major axis ) , the surface brightness changes more rapidly along the @xmath37 direction than along the @xmath57-direction
hence , the observed intensities constrain the @xmath58 more tightly than the @xmath56 . other model assumptions to fit the observed intensities
were also studied . in all cases the results for @xmath58 agreed to within a few mas , while the @xmath56 could differ by as much as @xmath59 .
most of the necessary data reduction steps are performed by the hst calibration ` pipeline ' . for the m32 data ,
three issues required additional attention : ( i ) wavelength calibration ; ( ii ) absolute sensitivity calibration ; and ( iii ) flat - fielding .
the wavelength scale provided by the calibration pipeline is not accurate enough for our project .
arc lamp spectra were therefore obtained in each orbit during occultation .
the emission line centers were determined , yielding for each arc spectrum @xmath60 and for each line @xmath61 with vacuum wavelength @xmath62 , the observed wavelength @xmath63 on the scale provided by the calibration pipeline .
subsequently , the observed offsets @xmath64 were fit as @xmath65 the constant offset @xmath66 is the result of non - repeatability in the fos grating and aperture wheels , and was measured to be @xmath67 .
the third order polynomial @xmath68 with zero mean accounts for a slight ( @xmath69 ) non - linearity of the wavelength scale .
the offsets @xmath70 account for constant shifts of the wavelength scale from orbit to orbit , mostly due to the ` geo - magnetically induced image motion problem ' ( keyes 1995 ) .
these shifts ranged between @xmath71 , and could be determined with @xmath72 ( @xmath73 ) accuracy .
the wavelength scale of the m32 spectra was corrected using the above fits to the arc spectra .
an additional shift of @xmath74 was applied to correct for the fact that the light path for external targets differs from that for the internal arc lamp ( keyes 1995 ) . for each m32 spectrum
the @xmath70 were used that was / were determined from the arc spectrum / spectra obtained immediately preceding or following the m32 spectrum .
the fact that the arc spectra are never obtained completely simultaneously with galaxy spectra introduces a slight additional uncertainty . the cumulative uncertainty in the mean streaming velocities of the stars as a result of uncertainties in the wavelength scale is estimated to be @xmath75 .
this is always smaller than the formal errors in these velocities as a result of photon noise ( cf .
table [ t : kinresults ] below ) .
after wavelength calibration ( and absolute sensitivity calibration and flat - fielding as described below ) the galaxy spectra were rebinned logarithmically in wavelength , as required for the stellar kinematical analysis .
a scale of @xmath76pixel was used .
the fos is a photon counting detector , so the formal errors in the spectra follow directly from poisson statistics . on average ,
the ( logarithmic ) rebinning decreases the formal errors , since it reduces the scatter between neighboring pixels .
the decrease can be calculated for each pixel , and was taken into account in the subsequent analysis .
the correlation between the errors in neighboring pixels induced by the rebinning was neglected .
the calibration pipeline multiplies the observed count rate spectra by a so - called ` inverse sensitivity file ' ( ivs ) , to obtain fluxes in @xmath77@xmath78 .
the pipeline ivs files contain a wavelength dependent aperture throughput correction , based on calibration observations of point sources .
these , however , are not applicable to extended sources .
therefore , instead of the standard pipeline ivs file a more recent ivs file with no aperture throughput correction was used , provided to us by fos instrument scientist tony keyes .
each wavelength in the m32 spectra samples a slightly different region of the galaxy , because of the wavelength dependence of the psf .
this influences only the continuum slope of the spectra , because the psf varies very slowly with wavelength . in the stellar kinematical analysis one
is only interested in the absorption lines , which are not influenced .
the underlying continuum is subtracted .
hence , no attempt was made to construct aperture throughput corrections .
( in fact , this would require accurate knowledge of the wavelength dependence of the psf , which is not readily available for the fos ; see appendix [ s : appa ] ) .
a g570h flat - field based on multiple 0.25-pair ( upper ) aperture observations of a star was used , as provided to us by tony keyes .
no flat - fields obtained explicitly with the 0.1-pair ( upper ) aperture were available .
a star illuminates the photocathode of the detector differently than an extended source . to test the appropriateness of the flat - field for the m32 data it was cross - correlated with the continuum subtracted normalized galaxy spectra .
this yielded clear peaks , indicating that the flat - field and the m32 spectra share the same features ( as they should ) .
the peak was only @xmath7910% lower for the 0.1-pair spectra than for the 0.25-pair spectra .
this indicates that the flat - field can be properly used for the 0.1-pair ( upper ) aperture as well , although a flat - field obtained specifically with that aperture would have been preferable .
shifts between the galaxy spectra and flat - field were identified from the positions of the cross - correlation peaks , and were corrected for ( see also kormendy 1996a ) .
the flat - fielding removes most of the pixel - to - pixel sensitivity variations , but some small residual variations might remain .
the observed spectrum of a source is the convolution of its actual spectrum with the line - spread - function @xmath80 .
the lsf can be written as the convolution of an ` illumination function ' @xmath81 and an ` instrumental broadening function ' @xmath82 ( cf .
appendix [ s : appb ] ) .
the function @xmath81 is the normalized intensity distribution of the light that falls onto the grating .
it depends on the target brightness and choice of aperture .
the direction of dispersion is parallel to the @xmath57 direction .
hence , @xmath81 is the integral over the @xmath37 direction of the two - dimensional brightness distribution that falls onto the grating . for the psf and aperture sizes derived in appendix [ s : appa ] , @xmath81 is given by equations ( [ blamdef ] ) and ( [ bkernel ] ) in appendix [ s : appb ] .
it can be calculated for each observation upon substitution of the lauer ( 1992 ) cusp model for the m32 surface brightness , and the aperture positions in table [ t : applaces ] .
the normalized function @xmath82 accounts for the instrumental broadening due to the grating and finite size of a detector diode ( the resolution element ) .
it was determined empirically from fits to the emission line shapes in the arc spectra , as discussed in appendix [ s : appb ] .
the lsf has zero mean if the galaxy light is distributed symmetrically within the aperture .
this is not generally the case , because the aperture centers are not exactly at @xmath83 ( figure [ f : aper_plot ] ) .
calculations show that the intensity weighted mean position is within @xmath84 from the aperture center for all the observations . nonetheless , there are noticeable wavelength shifts , because @xmath84 projects onto @xmath85 in the wavelength direction ( @xmath86 at 5170 )
. table [ t : applaces ] lists the lsf mean as calculated for each of the m32 spectra , both in and in @xmath6 at 5170 .
errors in these values due to errors in the aperture positions do not exceed 0.04 ( @xmath87 at 5170 ) .
the width and shape of the lsf are determined mainly by the aperture size .
the calculated lsfs for all the m32 spectra are shown in figure [ f : lsf ] .
the lsf shapes for the 0.1-pair observations are very similar to each other , as are the lsf shapes for the 0.25-pair observations .
the shapes are also similar to the emission line shapes observed directly in arc spectra ( figure [ f : arc ] ) .
the 0.25-pair lsf can be well approximated by convolving the 0.1-pair lsf by a gaussian with a dispersion of @xmath88 ( @xmath89 at 5170 ) , as illustrated by the dotted curve in figure [ f : lsf ] .
the best - fitting gaussian to the 0.1-pair lsf has a dispersion of @xmath90 ( @xmath91 at 5170 ) . for the 0.25-pair lsf
it has a dispersion of @xmath92 ( @xmath93 at 5170 ) .
however , these numbers are of limited use in characterizing the lsfs , since these have noticeably broader wings than a gaussian .
for stellar kinematical analysis a template spectrum is needed to compare the m32 spectra to . to avoid systematic errors
it is important to minimize template mismatching .
this is best done by constructing a composite template which contains an appropriate mix of spectral types .
observing template stars with the hst is inefficient and time consuming ( mainly because of the lengthy target acquisitions that are required ) . to date less than a handful different template stars have been observed with the hst , all of similar spectral type .
these hst spectra are insufficient to construct an optimal template .
a ground - based template library of spectra of 27 stars of different spectral types was therefore used , obtained in february 1990 by m. franx at the 4 m telescope of the kpno with the rc spectrograph ( as discussed previously by van der marel & franx 1993 ) .
the lsf of these spectra is approximately gaussian with a dispersion of @xmath94 ( @xmath95 at 5170 ) .
the spectra cover the spectral range from 4836 to 5547 , centered on the mg b triplet at @xmath96 .
this is the most useful wavelength range for stellar kinematical analysis , so it is no drawback that this range is smaller than the full range covered by the fos g570h grating .
the stellar spectra were similarly rebinned logarithmically as the galaxy spectra , and shifted to a common velocity .
no attempt was made to construct a different composite template for each m32 spectrum ( which is reasonable , given that there are no strong color gradients in the central arcsec ; lugger 1992 ) .
instead , the best fit was sought to a grand - total m32 spectrum , constructed by summing the hst spectra .
the resulting composite template is a weighted mix of the individual templates , with the weights determined using the method outlined in appendix a.3 of van der marel ( 1994a ) .
the mix contains giants , sub - giants and dwarfs of spectral types g and k. the composite template is used in the remainder of the paper .
however , other templates were studied as well .
for example , a stellar kinematical analysis was performed with a k - star template spectrum obtained from the hst data archive , taken with the fos circular 0.3 aperture ( @xmath97 diameter ) by h. ford and collaborators before the installation of costar .
this template provides a poor fit to the spectrum of m32 .
nonetheless , the stellar kinematical properties inferred with this template were found to be consistent with those derived using the composite template .
the stellar kinematical analysis was performed with the method presented in van der marel ( 1994a ) .
it fits the convolution of a parametrized velocity profile and a template spectrum to a galaxy spectrum in pixel space , using chi - squared minimization .
the formal errors in the fit parameters follow from the shape of the chi - squared surface near its minimum .
the method has been well tested , and its results agree with those from other methods for extracting stellar kinematics from galaxy spectra .
deviations of the line - of - sight velocity profiles from gaussians contain useful information on the dynamical structure of galaxies .
unfortunately , the velocity resolution of the fos is too poor to extract any reliable velocity profile shape information from the m32 data .
the analysis was therefore restricted to gaussian velocity profile fits .
data obtained with the fos is time resolved .
the red side detector reads and stores the data every @xmath98 minutes .
between 6 and 20 spectra were therefore available per aperture position .
the aperture positions for the observations l4 and l5 differ by @xmath99 , but both are at a distance of @xmath20 from the galaxy center , where kinematical gradients are small . in the kinematical analysis l4 and l5
were therefore treated as a single observation , referred to as l45 . especially for the 0.1-pair observations , the signal - to - noise ratio ( s / n ) of the individual read - outs is low .
individual read - outs for a given aperture position were therefore added to achieve an average s / n of @xmath100 per ( logarithmically rebinned ) pixel .
this resulted in 4 independent spectra for each of the observations s1s4 with the 0.1-pair aperture , and between 9 and 11 independent spectra for the observations l1 , l2 , l3 and l45 with the 0.25-pair aperture .
a stellar kinematical analysis was performed on each of these spectra .
the kinematical results for each aperture position were then averaged together , weighted with the errors .
it was found that this yields slightly more accurate results than a kinematical analysis of the sum of the individual spectra for a given aperture position .
no systematic trends with time in the orbit were found in the kinematical results .
the stellar kinematical analysis was performed over the wavelength range 48595520 .
regions influenced by so - called ` noisy ' diodes were masked .
the fit to the galaxy spectrum includes a low order polynomial to account for continuum differences between the galaxy and template spectrum .
this polynomial is fit simultaneously with the velocity profile .
the parameters of the best - fitting velocity profile are virtually independent of the choice of polynomial order .
a polynomial of order 5 was used . the mean velocity and velocity dispersion of the best - fitting gaussian velocity profiles
must be corrected for instrumental effects .
the mean velocity is biased because the lsf mean is non - zero .
this bias is removed by subtracting the corrections @xmath101 listed in table [ t : applaces ] .
these were calculated for the wavelength 5170 of the mg b triplet , but the wavelength dependence of the velocity corrections over the fit range is negligible ( @xmath102 in absolute value ) .
further , to obtain velocities in the m32 frame one must subtract a constant offset , determined by the difference between the systemic velocity of m32 and the template velocity .
these were not known accurately enough to determine the offset directly , and it was therefore estimated from the data itself .
somewhat arbitrarily , it was fixed to the intercept of the best least - squares fit line in a plot ( see figure [ f : allres]c below ) of rotation versus position @xmath37 along the m32 major axis for the observations s4 , s1 and s2 ( for which @xmath103 , @xmath104 and @xmath105 , respectively ) .
the velocity dispersions @xmath106 of the best - fitting gaussians must be corrected for differences in the lsfs of the template and galaxy spectra . for this
it is useful to consider not the observed dispersions themselves , but rather differences in the observed dispersions between two observations .
let @xmath107 and @xmath108 be the best - fitting dispersions for observations at positions a and b. assume that the lsf of observation b can be obtained from the lsf of observation a through convolution with a gaussian of dispersion @xmath109 .
the difference between the actual dispersions at positions a and b can then be estimated as : @xmath110 ( which uses the fact that the convolution of two gaussians is again a gaussian , the dispersion of which is the rms sum of the input dispersions ) .
this approach has two important advantages .
first , the lsf of the template influences the kinematical analysis at points a and b in the same way , so it does not enter into @xmath111 .
second , the differences in the lsfs for the hst observations can in fact to good approximation be accounted for by convolution with a gaussian ( cf .
figure [ f : lsf ] ) .
one does _ not _ have to assume that the lsfs themselves are gaussian , which would be a poor approximation . in the following , the observation l45 at @xmath20 from the galaxy center is used as ` reference point ' .
equation ( [ sigdef ] ) then yields @xmath112 where @xmath113 for an observation with the 0.25-pair aperture , and @xmath114 for an observation with the 0.1-pair ( based on the results of section [ s : lsf ] ) .
the latter value was calculated for the wavelength 5170 of the mg b triplet , which is roughly at the center of the fit range .
the actual value of @xmath115 varies from @xmath116 to @xmath117 over the fit range , which can be taken into account by assigning a ` formal ' error of @xmath118 to @xmath109 .
equation ( [ sigcor ] ) can be used to estimate the stellar velocity dispersion @xmath119 for any of the observations s1s4 and l1l3 , if one assumes that the dispersion @xmath120 at the position of observation l45 is known a priori . at @xmath121 along the major axis one
is beyond the region most influenced by seeing in ground - based data .
based on the data of van der marel ( 1994a ) and bender , kormendy & dehnen ( 1996 ) it was assumed that @xmath122 ( see figure [ f : allres]d below ) .
it is demonstrated in appendix [ s : appc ] that the velocity dispersions thus obtained have no systematic biases , and that even velocity dispersions as small as @xmath23 can be measured reliably , in spite of the large fos instrumental dispersion of @xmath123 .
the formal errors in dispersions @xmath119 obtained from equation ( [ sigcor ] ) follow from the errors in @xmath106 , @xmath124 , @xmath125 and @xmath126 using a standard first order error analysis .
systematic errors due to template mismatching are not likely to exceed a few @xmath6 , and are smaller than the formal errors in the measurements .
table [ t : kinresults ] lists the line strengths , rotation velocities and velocity dispersions , after correction for all instrumental effects .
these results are discussed below .
figure [ f : allres ] displays the results as function of the position @xmath37 of the aperture center along the m32 major axis .
the best spatial resolution ground - based m32 data are also shown .
these are the data from van der marel ( 1994a ) obtained at the william herschel telescope ( wht ) , and those of bender , kormendy & dehnen ( 1996 ) obtained at the canada france hawaii telescope ( cfht ) .
a comparison of these ground - based data to older ( lower spatial resolution ) ground - based data can be found in kormendy & richstone ( 1995 ) .
the spatial resolution for each of the observations in figure [ f : allres ] is characterized by a convolution kernel , which results from the psf and finite aperture size ( for ground - based long - slit data the effective aperture is the spatial ccd pixel size by the slit width ) .
the top left - panel illustrates the difference in spatial resolution between the data sets .
it displays the projection of the convolution kernel along the m32 minor axis , i.e. , the probability @xmath127 that a photon observed in a given aperture was emitted at distance @xmath37 from the center of that aperture , measured along the m32 major axis .
these probability distributions have fwhm values of @xmath128 , @xmath129 , @xmath130 and @xmath131 , for the hst / fos/0.1-pair , hst / fos/0.25-pair , cfht and wht data , respectively .
the hst data therefore provide a factor 5 increase in spatial resolution over even the best available ground - based data .
the line strengths shown in figure [ f : allres]b were normalized to an average of unity ( line strengths are measured with respect to the template spectrum , and are hence on an arbitrary scale ) .
there is no obvious trend with radius .
the observations with the 0.1-pair aperture closest to the nucleus have relatively low line strength , but this is unlikely to be real .
metallicities and line strengths generally increase towards the centers of galaxies
. the observed low values could be an artifact due to the presence of a black hole .
the stars that move rapidly near the black hole produce broad velocity profile wings .
these can be mistaken for an enhanced continuum , which causes the line strength to be underestimated ( van der marel 1994b ) .
the broad - band colors of m32 are constant in the central arcsec ( lugger 1992 ) . the true line strengths in the central @xmath132 covered by the hst observations are therefore probably also close to constant .
the observed line strengths have a sizable rms scatter of @xmath133 .
this exceeds the formal errors , which are on average @xmath134 .
the measurements are therefore not statistically consistent with a constant line strength ( as is confirmed by a chi - squared test ) .
the scatter in the observed line strengths could be the result of minor inaccuracies in the data , such as unidentified bad pixels or flat - field errors , or it could be real , and due to shot noise in the contributions from individual stars .
the total v - band luminosity inside the aperture varies from @xmath135 ( for observation s3 ) to @xmath136 ( for observation l3 ) .
most of the observed light comes from giants , some of which can be as luminous as @xmath137 ( freedman 1989 ) .
the luminosity of a ` typical ' giant as measured from surface brightness fluctuations ( tonry , ajhar & luppino 1990 ) is @xmath138 .
this implies that there are @xmath139 typical stars in each hst aperture .
therefore , @xmath140 fluctuations can be of order a few percent of the intrinsic line strength differences between stars .
figure [ f : allres]c shows the rotation velocities .
the hst rotation curve is smooth .
it is significantly steeper in the central @xmath141 than seen in the ground - based data , consistent with the presence of a central black hole .
the rotation velocity is observed to be @xmath142 at @xmath1 from the nucleus .
this is roughly twice the value measured from the ground at this distance .
the scatter in the line strength measurements ( figure [ f : allres]b ) induces scatter in the velocity dispersion measurements , because the velocity dispersion and line strength of a gaussian fit are statistically correlated .
tests indicate that a line strength error @xmath143 induces an error @xmath144 in the velocity dispersion .
rotation velocity measurements are not influenced .
to avoid unwanted errors in the velocity dispersion measurements , they were determined by fitting gaussian velocity profiles to the data while keeping the line strength constant to the mean of the measurements in figure [ f : allres]b .
this yields velocity dispersions that differ only at the @xmath145 level from those obtained by gaussian fits with variable line strength , so the main results do not depend on this correction .
figure [ f : allres]d shows the resulting velocity dispersions .
the dispersion measured with the 0.1-pair aperture closest to the nucleus is @xmath3 .
the profile of the velocity dispersion with radius is smooth , with the exception of the measurement @xmath146 from observation s2 , which is significantly lower than that for the other observations in the central @xmath1 .
we doubt the reality of this , but have not been able to attribute it to any known instrumental effect .
the average of the four dispersion measurements inside @xmath1 is @xmath4 , with a rms scatter of @xmath5 .
hence , the central velocity dispersion is definitely higher than the 8595@xmath6 determined from ground - based measurements .
it thus appears that a keplerian increase in the velocity dispersion close to the center of m32 has been resolved .
the difference in the dispersions measured for observations s1 and s2 , which were obtained with the same aperture at similar positions , indicates that uncertainties in the velocity dispersions are dominated by systematic errors , rather than poisson noise .
systematic errors in the rotation velocity measurements are believed to be much smaller , primarily because these are less sensitive to the poor instrumental resolution of the fos .
either way , the data clearly imply that the dispersion in the central @xmath1 is higher than measured from ground - based data . whether the nuclear dispersion seen through a small aperture
is indeed as high as @xmath147 should await verification by observations with the future hst spectrograph stis .
hst / fos spectra were obtained through small apertures centered on the nuclear region and major axis of m32 .
kinematical analysis of these data is more complicated than it is for ground - based data : the fos is a low - resolution spectrograph , and is not particularly well suited for dynamical studies of low - dispersion objects such as m32 .
one of the main results of this paper is therefore that stellar kinematical quantities _ can _ in fact be determined from the data , provided that a detailed study is made of various technical issues , including : calibration and modeling of the fos aperture sizes , psf and lsfs , determination of the aperture positions from the data , accurate wavelength calibration , template matching , and corrections for instrumental resolution .
the resulting kinematical quantities are displayed in figure [ f : allres ] .
the main result is that in the central @xmath1 the m32 rotation curve is significantly steeper , and the velocity dispersion significantly higher , than measured from even the highest resolution ground - based data .
these observational results are robust .
they do not depend critically on the corrections made in the analysis ; even a rough analysis without any template matching and corrections for instrumental effects shows these main features .
the results are exactly what would be expected if m32 does indeed have a massive dark object in its nucleus , as suggested previously on the basis of ground - based data . in the companion papers in this series ( van der marel 1997a , b ; cretton 1997 ) we construct dynamical models to address the presence of such a dark object quantitatively .
we also address the question whether the dark object has to be a black hole , or whether plausible alternatives still exist .
the authors are grateful to the fos instrument scientists tony keyes , anuradha koratkar and michael dahlem for helpful discussions , and to bill workman and jean surdej for successful scheduling and implementation of the observations .
tony keyes provided preliminary instrument calibration data and marijn franx provided the template library .
simon white helped with the early stages of the project .
support for this work was provided by nasa through grant number # go-05847.01 - 94a , and through a hubble fellowship # hf-1065.01 - 94a awarded to rpvdm , both from the space telescope science institute which is operated by the association of universities for research in astronomy , incorporated , under nasa contract nas5 - 26555 .
some calibrations exist of the hst+fos psfs and lsfs ( koratkar 1996 ) and of the fos aperture sizes ( evans 1995 ) , but these are too crude for the purposes of this paper . to obtain more accurate calibrations we present detailed models for the calibration observations obtained by evans ( 1995a ) after the installation of costar .
he centered a star in fos apertures , and stepped the star across each aperture in the fos @xmath57 and @xmath37 directions , respectively . at each step
the total intensity was measured . here
the results are considered for the fos red detector with three apertures : the square upper aperture of the 0.1-pair , the square upper aperture of the 0.25-pair , and the circular 1.0 .
the names of these apertures are based on their size in arcsec , _ before _ the installation of costar .
their nominal post - costar sizes are smaller by a factor @xmath148 .
figure [ f : transmis ] shows for each aperture the observed intensity as function of distance from the aperture center , normalized by the observed intensity with the star at the center of the circular 1.0 aperture .
the aperture length or diameter is easily determined from these data for apertures much larger than the psf core : it is then twice the radius at which the intensity has fallen to 50% of its central value .
this was used by evans ( 1995 ) to estimate the size of all fos apertures .
although this is reasonable for the larger apertures , it is not for the smaller apertures . for those a more careful analysis is required , in which the aperture sizes and psf are fitted to the data simultaneously . for a target with surface brightness @xmath149 ,
the intensity @xmath150 observed through an aperture centered on @xmath34 is @xmath151 where @xmath152 is a convolution kernel that depends on the psf and the aperture size and geometry .
the psf is assumed to be a circularly symmetric sum of gaussians , @xmath153 , \label{psfgauss}\ ] ] where the @xmath154 must satisfy @xmath155 . for a perfectly rectangular aperture with sides @xmath156 and
@xmath157 the convolution kernel is then @xmath158 -\mathop{\rm erf}\nolimits \left [ \fr{x-(a_x/2)}{\sqrt{2}\ > \sigma_i } \right ] \right \rbrace \
> \left \lbrace \mathop{\rm erf}\nolimits \left [ \fr{y+(a_y/2)}{\sqrt{2}\ > \sigma_i } \right ] -\mathop{\rm erf}\nolimits \left
[ \fr{y-(a_y/2)}{\sqrt{2}\ > \sigma_i } \right ] \right \rbrace , \ ] ] while for a perfectly circular aperture with diameter @xmath66 it is @xmath159 \int_0^{d/2 } \exp \left [ - \fr{1}{2 } ( \fr{r}{\sigma_i } ) ^2 \right ] \
> i_0(\fr{r \rho}{\sigma_i^2 } ) \
> r \ > dr , \ ] ] ( van der marel 1995 ) .
the function @xmath160 is the error function and @xmath161 is a bessel function . this model was used to fit the data in figure [ f : transmis ] . for a point source at @xmath162 ,
the intensity observed through an aperture centered at the origin is simply @xmath163 .
the solid curves in the figure show the best chi - squared fit to the data for a psf that is a sum of three gaussians .
the aperture sizes were treated as free parameters .
the square apertures were assumed to have equal sizes in the @xmath57 and @xmath37 directions .
the model was forced to fit 2 additional constraints , obtained from theoretical modeling ( as quoted in keyes 1995 ) : ( i ) the throughput ( for a centered point source ) of the circular 1.0 aperture is @xmath164 times that of a rectangular aperture of size @xmath165 ; and ( ii ) the absolute throughput of the @xmath165 aperture is @xmath166 .
the best - fitting model parameters are listed in table [ t : psfapfit ] .
the fit is robust , in spite of the fact that the model parameters are highly correlated . in particular : ( i ) no good fit can be obtained if the aperture sizes are held fixed to their nominal post - costar values ( which are @xmath29 square , @xmath30 square , and @xmath50 circular , respectively ) ; and ( ii ) the fit does not improve significantly if the number of gaussians in the psf is increased .
the formal errors in the best fit parameters are @xmath167 for the @xmath154 , and @xmath18% for the @xmath168 and aperture sizes .
these numbers are not too meaningful , however , because systematic errors probably exceed the formal errors .
figure [ f : psffit ] shows the psf for the best fit , and the encircled flux curve @xmath169 .
the latter is given by @xmath170 .
\label{encircled}\ ] ] the psf has 73% of the light within a circle of radius @xmath1 .
this is smaller than the 84% measured from star images at visual wavelengths obtained with the foc , but larger than the 65% measured with the wfpc2 ( which suffers from additional pixel scattering ) .
it should be noted that the psf derived here pertains only to the upper position of the square paired fos apertures .
the aperture transmissions , and thus the psf , are somewhat different at the lower positions of the square paired apertures ( evans 1995a ) .
the above analysis ignores the effects of diffraction at the aperture edges .
the derived aperture sizes are therefore not the geometrical apertures sizes .
this might explain why the derived sizes are smaller than the nominal post - costar sizes .
non - circularly symmetric features in the psf are also ignored .
such features might in fact be present , given that the results of scanning a star across the aperture in the @xmath57 and @xmath37 directions differ slightly ( figure [ f : transmis ] ) .
either way , both the aperture sizes and the psf enter into the data analysis only through the convolution kernels @xmath152 .
these kernels are adequately fit by the model , independent of whether or not the aperture sizes and psf of the model are unbiased estimates of their true values .
the observations in figure [ f : transmis ] were derived from so - called ` white light ' images , i.e. , with no grating in the light path .
the psf with the parameters of table [ t : psfapfit ] is therefore an average over all wavelengths to which the detector is sensitive .
the m32 observations were obtained with the g570h grating .
observations with this grating ( bohlin & colina 1995 ) show fractional throughputs for centered point sources that are larger by a few percent than those in figure [ f : transmis ] .
this indicates that the core of the psf for the g570h grating might be @xmath171% smaller than that of the psf derived here . on the other hand , in fitting the m32 peak - up acquisition observations ( section [ s : acq_results ] ) by convolving the lauer ( 1992 ) cusp model for the m32 surface brightness
, it was found that the chi - squared of the fit could be improved by _ increasing _ the width of the psf core by @xmath172% .
this could mean that the psf does in fact have a broader core than derived here , or alternatively , that the lauer model , based on wfpc1 measurements , is not a perfectly adequate representation of the m32 surface brightness in the central @xmath1 . in the absence of theoretical constraints on the wavelength dependence of the psf from accurate wavefront modeling ,
it was decided not to make any corrections to the psf derived from the white light images ( table [ t : psfapfit ] ) .
the psf remains somewhat uncertain , but it was verified that changes in the psf core width between @xmath173% and 20% do not change any of the major conclusions of our paper(s ) .
the observed spectrum @xmath174 of a source is the convolution of its actual spectrum @xmath175 with the line - spread - function @xmath80 , @xmath176 the lsf can be written as the convolution @xmath177 where the ` illumination function ' @xmath81 is the normalized intensity distribution of the light that falls onto the grating , and the normalized ` instrumental broadening function ' @xmath82 accounts for the broadening due to the grating and the detector .
the @xmath34 coordinate system is defined such that the @xmath57 direction is parallel to the direction of dispersion .
the function @xmath81 is therefore proportional to the integral over the @xmath37 direction of the two - dimensional brightness distribution that falls onto the grating .
all the light in the aperture is detected , since the apertures used for the observations are smaller than the @xmath37-size of the diode array of the detector ( @xmath178 ) . for the case of observations of a target with surface brightness @xmath149 through an aperture centered on position @xmath162 , @xmath179 where @xmath180 is defined in equation ( [ itot ] ) , and @xmath181 is a convolution kernel that depends on the psf and the aperture size and geometry . for the g570h grating the wavelength @xmath182
is related to the position @xmath57 according to @xmath183(x / arcsec ) .
if the aperture is rectangular of size @xmath184 , and the psf is as given in equation ( [ psfgauss ] ) , then @xmath185 \ > \left \lbrace \mathop{\rm erf}\nolimits \left [ { { y'+(a_y/2 ) } \over { \sqrt{2}\ > \sigma_i } } \right ] -\mathop{\rm erf}\nolimits \left [ { { y'-(a_y/2 ) } \over { \sqrt{2}\ > \sigma_i } } \right ] \right \rbrace , & \quad \vert x \vert \leq { { a_x}\over2 } ; \\ 0 ,
& \quad \vert x \vert > { { a_x}\over2 } .
\end{array } \right .
\label{bkernel}\ ] ] the grating disperses the image of the aperture onto the photocathode of the detector , which is scanned using an array of diodes .
a simple model for the instrumental broadening function @xmath82 is therefore : @xmath186 where @xmath187 is the normalized response function of a detector diode and @xmath188 is the normalized broadening function of the grating .
let @xmath187 be a top - hat function with full width @xmath189 , and let the grating broadening function be @xmath190 , where @xmath191 is a normalized function , and @xmath192 is a free parameter that measures the width of @xmath193 .
the notation @xmath194 is used for the primitives of @xmath191 .
the function @xmath195 is then @xmath196 .\ ] ] it can be determined empirically by fitting to the emission lines in the arc spectra . if the arc lamp illuminates the aperture homogeneously , then @xmath81 is a normalized top - hat function of full width @xmath156 . for an emission line that is intrinsically a delta - function at @xmath197 , the observed line shape will be @xmath198 .
\label{oarc}\ ] ] several functional forms were considered for @xmath191 .
a gaussian did not yield good fits to the observed line shapes .
satisfactory results were obtained with a lorentzian , for which @xmath199 .\ ] ] with this choice , the function @xmath195 has two parameters , @xmath192 and @xmath189 .
the best fit is generally attained for values of @xmath189 that differ from the actual size of one diode , which is @xmath200 for the g570h grating .
this probably indicates that the true diode response function is not a perfect top - hat .
however , this does not invalidate the model .
one is interested only in a fit to the function @xmath82 , not in a determination of either @xmath187 or @xmath188 .
therefore , both @xmath192 and @xmath189 are treated as free fitting parameters .
the best fits were determined to the emission line shapes in arc spectra obtained with the 0.1-pair and the 0.25-pair apertures , while keeping the values of @xmath156 fixed to their values in table [ t : psfapfit ] .
figure [ f : arc ] shows the fits to the arc lines and the best - fitting function @xmath195 .
no evidence was found for a wavelength dependence of the lsf over the wavelength range covered by the grating .
the best - fit parameters are : @xmath201 and @xmath202 for the 0.1-pair aperture , and @xmath203 and @xmath204 for the 0.25-pair aperture . hence , the function @xmath82 is broader for the 0.25-pair aperture than for the 0.1-pair .
the difference is too large to be attributed to small errors in the assumed aperture sizes .
hence , the instrumental broadening is more complicated in reality than in our simplified model , in which one would have expected @xmath82 to be independent of the aperture size .
however , the only important point for the interpretation of the m32 spectra is that an accurate empirical description of the lsf is available .
the fits in figure [ f : arc ] suggest that the model is satisfactory for this purpose .
tests with artificial galaxy spectra were performed to determine whether systematic biases could be present in the velocity dispersion measurements .
artificial galaxy spectra were created from each of the 27 stellar spectra in the template library , by broadening them with gaussian velocity profiles with dispersions @xmath205 .
twenty copies were made of each broadened template , and noise was added to yield a s / n of 10 per pixel .
a stellar kinematical analysis was then performed on each copy , using the composite spectrum discussed in section [ s : temp ] as template .
the twenty velocity dispersion measurements where then averaged to yield an output velocity dispersion @xmath206 .
figure [ f : kintest ] plots @xmath206 as function of @xmath205 for each of the input spectra .
the results lie closely along the line @xmath207 .
the minor deviations can be attributed entirely to the template mismatch in the simulations ( the input stellar spectra are giants , sub - giants and dwarfs of spectral types g , k and m ) . at dispersions below @xmath208
there is some tendency for the dispersion to be underestimated .
however , for the m32 observations one is always in the situation where the difference in dispersion between the galaxy and template spectrum is @xmath209 ( because the lsf of the galaxy spectra is broader than that of the template , in addition to the kinematical doppler broadening ) . at these dispersions
there are no significant biases in the results of the kinematical analysis , but only small ( @xmath210 ) errors due to template mismatching .
these errors are nearly independent of the velocity dispersion .
since the approach used in this paper for determining velocity dispersions relies on differences between velocity dispersion measurements at different positions , it is expected that even these small errors largely cancel out .
a crude consistency check on our approach for determining velocity dispersions can be obtained by making the ( poor ) assumption that the lsfs of both the galaxy and template spectra are gaussian , with dispersions @xmath211 and @xmath212 , respectively .
the stellar velocity dispersion @xmath119 then follows from the dispersion @xmath213 obtained by fitting to the galaxy spectrum , according to : @xmath214 for the observation l45 the dispersion measured directly from the spectra is @xmath215 . upon substitution in equation ( [ sigcorgau ] ) of @xmath216 ( cf .
section [ s : temp ] ) , and @xmath217 ( cf .
section [ s : lsf ] ) one obtains the estimate @xmath218 for the stellar velocity dispersion at @xmath20 along the m32 major axis .
this agrees reasonably well with the value of @xmath219 adopted in section [ s : correc ] on the basis of ground - based measurements .
this confirms the result of appendix [ s : ctests ] that there are no systematic biases in the velocity dispersion measurements from the spectra , and that the differences in the lsfs of the template and galaxy spectra can be fully corrected for .
evans , i. n. , koratkar , a. p. , keyes , c. d. , & taylor , c. j. 1995 , fos instrument science report
cal / fos-138 , ` smov report vii : fos aperture alignments ii .
small apertures and adopted alignments ' ( baltimore : space telescope science institute ) koratkar , a. , wheeler , t. , evans , i. , lupie , o. , taylor , c. , keyes , c. , & kinney , a. 1994 , fos instrument science report cal / fos-123 , ` smov report v : fos plate scale and orientation ' ( baltimore : space telescope science institute ) ccccccccr 1 & 4.3 & 3.66 & 1.29 & 2 & 7 & 3.36 & 0.80 & 4.7 + 2 & 1.0pair & 0.86 & 0.86 & 6 & 2 & 0.61 & 0.65 & 7.4 + 3 & 0.5pair & 0.43 & 0.43 & 3 & 3 & 0.29 & 0.29 & 9.1 + 4 & 0.25pair & 0.215 & 0.215 & 3 & 3 & 0.143 & 0.143 & 8.5 + 5 & 0.1pair & 0.086 & 0.086 & 5 & 5 & 0.043 & 0.043 & 38.7 + ccclcccc s1 & y2uf0107 t & 3 & 0.1-pair & @xmath220 & @xmath221 & @xmath222 & 2400 + s2 & y2uf0109 t & 4 & 0.1-pair & @xmath220 & @xmath221 & @xmath222 & 2400 + s3 & y2uf010bt & 5 & 0.1-pair & @xmath220 & @xmath221 & @xmath223 & 2400 + s4 & y2uf010dt & 6 & 0.1-pair & @xmath220 & @xmath221 & @xmath224 & 2400 + l1 & y2uf010gt & 7 & 0.25-pair & @xmath225 & @xmath221 & @xmath226 & 1105 + l2 & y2uf010ht & 7 & 0.25-pair & @xmath225 & @xmath221 & @xmath227 & 1105 + l3 & y2uf010jt & 8 & 0.25-pair & @xmath225 & @xmath221 & @xmath222 & 1395 + l4 & y2uf010kt & 8 & 0.25-pair & @xmath225 & @xmath221 & @xmath228 & 800 + l5 & y2uf010mt & 9 & 0.25-pair & @xmath225 & @xmath221 & @xmath228 & 1790 + llcccccc s1 & 0.1-pair & 0.068 & 218.1 & @xmath229 & @xmath230 & @xmath231 & @xmath232 + s2 & 0.1-pair & 0.068 & 203.6 & @xmath233 & @xmath234 & @xmath233 & @xmath235 + s3 & 0.1-pair & 0.068 & 139.7 & @xmath233 & @xmath236 & @xmath233 & @xmath235 + s4 & 0.1-pair & 0.068 & 199.2 & @xmath233 & @xmath237 & @xmath233 & @xmath235 + l1 & 0.25-pair & 0.191 & 686.0 & @xmath238 & @xmath239 & @xmath240 & @xmath241 + l2 & 0.25-pair & 0.191 & 1133.7 & @xmath238 & @xmath242 & @xmath243 & @xmath244 + l3 & 0.25-pair & 0.191 & 1269.3 & @xmath238 & @xmath245 & @xmath246 & @xmath247 + l4 & 0.25-pair & 0.191 & 435.5 & @xmath238 & @xmath248 & @xmath249 & @xmath250 + l5 & 0.25-pair & 0.191 & 427.7 & @xmath251 & @xmath252 & @xmath253 & @xmath254 + lrrrrrr s1 & 0.941 & 0.038 & @xmath255 & 10.0 & 156.4 & 10.2 + s2 & 0.957 & 0.033 & @xmath256 & 6.4 & 95.6 & 9.8 + s3 & 1.014 & 0.046 & @xmath257 & 9.7 & 111.9 & 11.7 + s4 & 1.049 & 0.039 & @xmath258 & 8.6 & 127.0 & 9.6 + l1 & 0.927 & 0.027 & @xmath259 & 5.3 & 78.3 & 11.0 + l2 & 0.989 & 0.022 & @xmath260 & 4.3 & 83.4 & 9.3 + l3 & 1.104 & 0.022 & @xmath261 & 4.8 & 124.2 & 7.0 + l45 & 0.973 & 0.022 & @xmath262 & 4.3 & 70.0@xmath12 & 5.0 + | we have obtained spectra through small apertures centered on the nuclear region and major axis of m32 , with the faint object spectrograph ( fos ) on the hubble space telescope ( hst ) . a detailed analysis and reduction of the data
is presented , including : ( i ) new calibrations and modeling of the fos aperture sizes , point - spread - function and line - spread - functions ; ( ii ) determination of the aperture positioning for each observation from the observed count rate ; and ( iii ) accurate wavelength calibration , template matching , and kinematical analysis of the spectra .
this yields measurements of the stellar rotation velocities and velocity dispersions near the center of m32 , with five times higher spatial resolution than the best available ground - based data .
the inferred velocities provide the highest angular resolution stellar kinematical data obtained to date for any stellar system .
the hst observations show a steeper rotation curve and higher central velocity dispersion than the ground - based data .
the rotation velocity is observed to be @xmath0 at @xmath1 from the nucleus .
this is roughly twice the value measured from the ground at this distance .
the nuclear dispersion measured through the smallest fos aperture ( @xmath2 square ) is @xmath3 .
the average of four independent dispersion measurements at various positions inside the central @xmath1 is @xmath4 , with a rms scatter of @xmath5 .
the nuclear dispersion measured from the ground is only 8595@xmath6 , whereas the dispersion outside the central arcsec is only @xmath7@xmath8 .
these results significantly strengthen previous arguments for the presence of a massive nuclear black hole in m32 .
detailed dynamical models are presented in a series of companion papers .
# 1#2 |
mechanical systems subject to rolling kinetic constraints are one of the most studied argument of classical mechanics , especially for its wideness of applicability in several branches of mechanical sciences : contact mechanics , tribology , wear , robotics , ball bearing theory and control theory applied to moving engines and vehicles are only some of the important fields where the results about pure rolling constraint can be fruitfully used .
it is well known that , when a mechanical system moves in contact with an assigned rough surface , the effective fulfilment of the kinetic conditions determined by the rolling without sliding requirement of the system on the surface depends on the behavior , with respect to the considered law of friction , of the reaction forces acting on the system in the contact points .
for example , the roll of a disk on a rough straight line , considering the coulomb s law of friction , can happen only if the contact force lie inside the friction cone ( see example 1 below ) .
however , even in the simplest case of a mechanical system formed by a single rigid body , in the case of multiple contact points between the rigid body and the rough surface , it could be an hard task to obtain sufficient information about the contact reactions in order to determine if the laws of friction are satisfied or not during the motion .
in fact the most common methods to determine information about the reactions , starting from the simple application of linear and angular momenta equations ( see e.g. @xcite ) to most refined techniques such as lagrangian multipliers in lagrangian mechanics ( see e.g. @xcite ) or deep analyses of the contact between the system and the surface ( see e.g. @xcite ) , have a global character .
then these methods , for their very nature , can determine only a reactive force system equivalent to the real one but , in the general case , these methods can not determine the single reactive forces in the contact points .
the problem becomes even more complicated in case of multibody system , due to the presence of the internal reactions in the link between the parts of the system . in this paper
we consider the motion of a mechanical system having two or more distinct contact points with one or more assigned rough surfaces , and we determine necessary conditions for which in all the contact points the pure rolling kinetic constraint can hold .
we also analyze the sufficiency of these conditions by generalizing to this case a well known and usually accepted assumption on the behavior of pure rolling constraint .
moreover , we briefly discuss the possible behaviors of the system when the necessary conditions are not fulfilled . the procedure to determine if the rolling condition can be fulfilled can be applied both to systems formed by a single rigid body and to multibody systems .
it is essentially based on the application of linear and angular momenta equations to the ( parts forming the ) mechanical system , and therefore it gives an underdetermined system in the unknown single contact reactions .
nevertheless , we show that the lack of complete knowledge of the single contact reactions is not an obstacle to determine the feasibility of the rolling conditions .
it is however important to remark that , although the procedure has a very simple and unassailable theoretic foundation , its effective application to general systems could present insurmountable difficulties .
this is essentially due to the fact that the general procedure explicitly requires the knowledge of the motion law of the system , and in the general case the explicit time
dependent expression of the motion can not be obtained because of complications determined by the geometry of the system itself and/or by the integrability of the equations of motion .
nevertheless there are several significative cases where the procedure can be explicitly performed . in the paper , we illustrate three examples with rising complication : the well known case of a disk falling in contact with an inclined plane ( that is presented only to point out some key points of the general procedure ) ; the case of a system formed by a non coupled pair of disks connected with a bar and moving on the horizontal plane ; the case of a heavy sphere falling in contact with a guide having the form of a v groove non symmetric with respect to the vertical axis and inclined with respect to the horizontal .
the main content of this paper can be approached starting from a very standard background knowledge , essentially focused to the linear and angular equations of motion for a mechanical system , the so called cardinal equations , and the basic theory of pure rolling conditions and kinetic constraints . on the other hand , the list of possible references involving theory and application of pure rolling constraint is almost endless .
therefore we chose to cite only a very limited list of references sufficient to make the paper self consistent : the classical book of levi civita and amaldi @xcite and the book of goldstein @xcite for the cardinal equations and the basic concepts about pure rolling conditions ; the book of neimark and fufaev @xcite and the paper of massa and pagani @xcite for the behavior of systems subject to kinetic constraints .
the interested reader can find in the wide but not exhaustive lists of references of @xcite as a useful starting point to delve in the expanse of the material related to this argument .
the paper is divided in four sections .
section 1 contains a very brief preliminary description of the well known analysis of the rolling condition for a disk in contact with an inclined plane .
this remind is motivated by some useful affinities with the general procedure for generic systems .
section 2 contains the discussion of the general case , and the determination of the necessary conditions for pure rolling conditions simultaneously hold .
section 3 presents the example of the system formed by the non coupled disks and the example of the heavy sphere falling in the v groove .
section 4 is devoted to open problems , remarks and conclusions .
an homogeneous disk of mass @xmath0 and radius @xmath1 moves in the vertical plane being in contact with a rough guide inclined with slope angle @xmath2 . [ disco ] considering the system subject to the coulomb s law of friction , with obvious notation clarified by fig .
[ disco ] , the feasibility of pure rolling condition of the disk can be determined with the following procedure : * we determine the relative velocity @xmath3 of the contact point @xmath4 of the disk at the instant @xmath5 with respect to the inclined plane as function of the initial data of the motion .
the pure rolling condition requires of course that @xmath6 .
if so * we assume that the disk rolls without sliding on the inclined plane .
then the system has a single degree of freedom ( for example the coordinate @xmath7 of @xmath4 along the inclined plane ) and we can determine the equation of motion @xmath8 * we determine the corresponding reaction @xmath9 as a ( in this case constant ) function of time @xmath10 * we test the coulomb s law of friction condition @xmath11 where @xmath12 are the parallel and orthogonal component of @xmath13 with respect to the inclined plane ; * we assume that , if and until the coulomb s condition is verified , the disk moves rolling on the plane and that if and when the coulomb s condition is not verified , the disk changes its dynamic evolution beginning to slide on the plane ( until the first time @xmath14 such that @xmath15 ) .
some remarks are in order to focus the possibility to generalize the procedure to more complicated systems .
step 2 consists in the determination of the reaction acting on the disk as function of time .
the utmost simplicity of the specific problem can hide the fact that in a more general situation the information about the reaction sufficient to analyze the rolling condition could require an explicit determination of the motion of the system as function of time .
step 3 tests the compatibility of the reaction evaluated in step 2 with the coulomb s law of friction assumed as the constitutive characterization of the rough surface in contact with the disk .
of course the feasibility of the rolling condition can be tested with any other significative constitutive law . in step 4
we assume that , roughly speaking , if the disk can roll then it does .
this is of course an arbitrary assumption , but the hypothesis is well confirmed by experimental results . in the next section , we will confirm this assumption in the more general situation of generic system . to conclude the section ,
let us note that , in this very simple case , both the behaviors of the disk when the coulomb s friction condition is or is not verified are determinable . in the general case ,
when the constitutive law is not verified , the behavior of the system turns out to be not so straight to determine , although some reasonable assumptions can be done .
we will go back on this arguments in section 4 .
in this section , following a line of though similar to the one applied in the previous section , we discuss the possibility that a mechanical system @xmath16 having two points @xmath17 in contact with a fixed surface @xmath18 moves such that in both the contact points the rolling conditions can subsist respecting the coulomb s law of friction .
the arguments of the discussion can be easily extended to cases with more ( but a finite number ) than two contact points and possibly to different friction constitutive laws .
the discussion is based on the fact that , along the motion , the reactive forces acting on the system must validate the linear and angular momenta equations @xmath19 where @xmath20 is the total mass of the system , @xmath21 is the center of mass of the system and @xmath22 are respectively the sum of the active and reactive forces and active and reactive momenta acting on the whole system . in this specific situation
we have that : @xmath23 it is however well known @xcite that eqs .
( [ sistema_generale ] ) are not sufficient to determine the motion of the mechanical system and the single reactions @xmath24 along the motion , since the system @xmath25 is by its very nature under determined .
in fact the projection of the angular momenta equation of ( [ sistema_sottodeterminato ] ) in the direction of @xmath26 is a pure equation of motion of the system where no reactions appear . then ( [ sistema_sottodeterminato ] ) can give no more than 5 relations on the components of @xmath27 and @xmath28 . unfortunately ,
due to the roughness of the contacts , no preliminary conditions can be imposed on the components of the reactions , so that , even when the motion of the mechanical system is known , ( [ sistema_sottodeterminato ] ) is a linear system with @xmath29 unknowns that is not of maximum rank .
nevertheless the parametric solution of the system ( [ determinazione_reazioni ] ) , and an assumption parallelizing the one of step 4 of the case of section 1 , give us the possibility of determining if the rolling conditions in @xmath30 and @xmath31 are or not verified .
the procedure to test the feasibility of pure rolling condition of the disk can be then based on the following steps : * we test if the initial relative velocities of the contact points @xmath17 with respect to the surface are null or not . if they are null * we suppose that the system rolls without sliding in both the contact points .
this assumption fixes the dynamics ( for example the number of degrees of freedom ... ) of the system and consequently allows the determination of the motion of the system ; * we write the linear and angular momenta equations for the whole system , for example in the form : @xmath32 since the motion of the system is known , both the right hand sides of the equations , together with the position vectors @xmath33 and @xmath34 , are known as function of time
. therefore eqs .
( [ sistema_reazioni ] ) turn out to be a time
dependent under determined linear system in the six scalar unknowns given by the components of the vectors @xmath35 ; * we solve the linear under determined system ( [ sistema_reazioni ] ) , obtaining the expression of the reaction @xmath27 and @xmath28 as function of time and parameters @xmath36 , where of course the integer @xmath37 is related to the rank of ( [ sistema_reazioni ] )
. then we can determine the tangent and orthogonal components @xmath38 of the reactions with respect to the surface @xmath18 as functions of @xmath39 .
the pure rolling conditions then can subsist in both the contact points only in the time interval @xmath40 $ ] such that for every @xmath41 $ ] there exists at least one admissible @xmath37uple @xmath42 such that the system @xmath43 holds ; * we assume that , if for every @xmath41 $ ] there exists at least one admissible @xmath37uple @xmath42 such that ( [ condizioni_cm ] ) are verified , the system moves rolling without sliding in both points @xmath30 and @xmath31 during the time interval @xmath40 $ ] .
it is clear that the general procedure described above parallelizes as possible and generalizes the one of the disk on the inclined plane .
the most significant differences consist in the explicit determination of the motion of the system ( since otherwise eqs .
( [ sistema_reazioni ] ) could not admit a simple parametric solution for the reactions @xmath44 and @xmath28 ) and in the fact that , when the coulomb conditions ( [ condizioni_cm ] ) are not verified , being understood that the system does not roll in both contact points , the determination of the behavior of the system could require a more subtle analysis .
we will go back on these arguments in section 4 .
we also remark that not all the @xmath37uple @xmath36 could be admissible in the discussion of the inequalities ( [ condizioni_cm ] ) .
for example , if the system is leaned on the surface , we have to restrict our attention to the @xmath37uple such that @xmath45 ( where @xmath46 is the unit normal vector to the surface @xmath18 in the point @xmath47 and orientated toward the side of the system ) since otherwise the system detaches from the surface .
[ ruote ] a mechanical system is formed by two equal disks of mass @xmath0 and radius @xmath1 and a rod , of mass @xmath20 and length @xmath48 . the rod is constrained to remain orthogonal to the two planes of the disks with its endpoints coinciding with the two centers of the disks ( see fig .
[ ruote ] ) so that the disks remain vertical .
the whole system is leaned on a rough horizontal plane .
the system has then @xmath49 degrees of freedom : the coordinates @xmath50 of the center of mass @xmath21 of the rod , the angle @xmath51 formed by the plane of the disks with the @xmath52 plane and the two rotation angles @xmath53 of the disks .
the rolling conditions in the contact points @xmath30 and @xmath31 are equivalently expressed by : @xmath54 tedious but straightforward computations ( see @xcite ) give the equations of motion of the system @xmath55 if we suppose assigned the almost generic initial data @xmath56 with the only condition @xmath57 , the motion of the system is given by @xmath58 + x_0 \\ \\ { y}(t ) & = & - \dfrac12 l \dfrac{\dot{\vph_1}_0 + \dot{\vph_2}_0}{\dot{\vph_1}_0 - \dot{\vph_2}_0}\left[\cos\left(\dfrac{r}{l } \left(\dot{\vph_1}_0 - \dot{\vph_2}_0\right)t + \vth_0 \right ) - \cos \vth_0 \right ] + y_0 \\
\\ \vth ( t ) & = & \dfrac{r}{l } ( \dot{\vph_1}_0 - \dot{\vph_2}_0)t \\ \\ { \vph_1}(t ) & = & \dot{\vph_1}_0 t \\ \\ { \vph_2}(t ) & = & \dot{\vph_2}_0 t \end{array } \right.\end{aligned}\ ] ] the linear and angular momenta equations for the system can be written as @xmath59 taking into account the motion of the system ( [ moto_es1 ] ) and introducing the orthonormal base @xmath60 with @xmath61 , with obvious notation we obtain @xmath62 \\ \\
\phi_{2_z } & = & \dfrac12 \left[(2m+m)g - ( 3m+m ) \dfrac{r^3}{l^2}(\dot{\vph_1}^2_0 - \dot{\vph_2}^2_0 ) \right ] \\ \\ \phi_{1_v } & = & 0 \\ \\ \phi_{2_v } & = & 0 \\ \\ \phi_{1_u } + \phi_{2_u } & = & - \dfrac12 ( 2m+m ) \dfrac{r^2}{l } ( \dot{\vph_1}^2_0 - \dot{\vph_2}^2_0 ) \end{array } \right.\end{aligned}\ ] ] note that , if the system leans on the horizontal plane , we must add the requirement @xmath63 since otherwise one between @xmath64 and @xmath65 becomes negative ( and this is not acceptable , because the system lifts from the horizontal plane , and the initial assumptions of five degrees of freedom is violated ) .
if ( [ condizioni_esistenza_es1 ] ) is fulfilled , then we can chose for example @xmath66 and we find the reactions @xmath24 as functions of @xmath67 : coulomb conditions ( [ condizioni_cm ] ) then takes the form : @xmath68 \\ \\
\left| \dfrac12 ( 2m+m ) \dfrac{r^2}{l } ( \dot{\vph_1}^2_0 - \dot{\vph_2}^2_0 ) + \lambda \right| \le \mu_2 \ , \dfrac12 \left[(2m+m)g - ( 3m+m ) \dfrac{r^3}{l^2}(\dot{\vph_1}^2_0 - \dot{\vph_2}^2_0 ) \right ] \end{array } \right.\end{aligned}\ ] ] in conclusion , the pure rolling of the disks can subsist if and only if ( [ condizioni_esistenza_es1 ] ) holds and there is a @xmath67 such that @xmath69 , \right .
\\ \qquad\qquad\left .
- \dfrac12 ( 2m+m ) \dfrac{r^2}{l } ( \dot{\vph_1}^2_0 - \dot{\vph_2}^2_0 ) \right .
\\ \qquad\qquad\qquad\qquad\left .
- \dfrac12 \mu_2 \left[(2m+m)g - ( 3m+m ) \dfrac{r^3}{l^2}(\dot{\vph_1}^2_0 - \dot{\vph_2}^2_0 ) \right ] \right\ } \\ \\
\qquad\qquad\qquad\qquad\qquad\qquad \le \lambda \le \\
\\ \qquad\qquad \min \left\ { \dfrac12 \mu_1 \left[(2m+m)g - ( 3m+m ) \dfrac{r^3}{l^2}(\dot{\vph_2}^2_0 - \dot{\vph_1}^2_0 ) \right ] , \right .
\qquad\qquad \qquad\qquad \dfrac12 \mu_2 \left[(2m+m)g - ( 3m+m ) \dfrac{r^3}{l^2}(\dot{\vph_1}^2_0 - \dot{\vph_2}^2_0 ) \right ] \right . \\ \qquad\qquad
. - \dfrac12 ( 2m+m ) \dfrac{r^2}{l } ( \dot{\vph_1}^2_0 - \dot{\vph_2}^2_0 ) \right\}. \end{array}\end{aligned}\ ] ] [ sfera ] a mechanical system is formed by a sphere of mass @xmath0 and radius @xmath1 leaned in an inclined v
groove whose walls are described by the equations @xmath70 we introduce an orthonormal base @xmath71 where @xmath72 are orthogonal to @xmath73 respectively and @xmath74 . the center @xmath75 of the sphere
is then determined by the vector @xmath76 , where @xmath7 is the distance of @xmath75 from a fixed plane orthogonal to @xmath77 ( see fig .
[ sfera ] ) .
the rolling conditions in the contact points @xmath78 determines the angular velocity of the sphere in the form @xmath79 and the system has one degree of freedom : the coordinate @xmath7 . the linear and angular momenta equations for the system can be written as @xmath80 the projection of the angular momenta equation in the direction of @xmath81 gives the equation of motion of the sphere , that is @xmath82 .
this relation suffices to obtain from ( [ sistema_generale_es2 ] ) the under determined system of the reactions : if we decompose the reactions along the basis introduced above @xmath83 the system takes the form @xmath84 to analyze the parametric solution of the system we chose @xmath85 . in this case , and once again supposing the sphere leaned on the groove , we must require the condition @xmath86 since otherwise @xmath87 and the sphere comes off the groove . conditions ( [ condizioni_cm ] ) take in this case the form @xmath88 with @xmath89 .
a straightforward minimum computation for the functions on the left - hand side of ( [ condizioni_cm_es2 ] ) shows then that the system can roll on both the contact points if and only if @xmath90
the procedure described in sec .
2 in the case of two contact points can be generalized to ( multibody ) systems with three or more contact points ( think for example of a steering tricycle formed by three vertical disks connected with three rods leaned on the horizontal plane ) . of course , an increase of the number of contact points implies in general an increase of the technical difficulties in practical applications .
this is principally due to the fact that step 1 of the general procedure is not a straightforward passage .
the effective knowledge of the motion of the system can be achieved only in some particular cases .
insurmountable technical difficulties can arise both for geometrical reasons ( think of a convex rigid body moving in contact with a surface , both having generic shapes with the only requirement that the contact between rigid body and groove happens in two points . for a more detailed discussion on the argument , see , e.g. @xcite ) , and/or for computational reasons ( even when the equations of motion of the system are explicitly obtained , it could be hard to integrate them to obtain the motion of the system ) .
nevertheless note that , as pointed out by the examples in sec .
3 , not for all the systems the explicit integration of the equations of motion is required .
a second remark is that the general procedure gives necessary conditions such that the pure rolling subsists in all contact points ( conditions that become sufficient if we take into account step 4 of the procedure ) but it does not give any information on the behavior of the system if the pure rolling is not possible even in a single contact point .
in fact , analogously to what happens in the simple case of ex . 1 , in the instant
when ( [ condizioni_cm ] ) stops to hold , the dynamics of the system ( for example , the number of degrees of freedom ) changes abruptly . to clarify this fact , suppose that , at the instant @xmath91 of the study of the system of ex .
2 , a sudden variation of the friction coefficient @xmath92 in the point @xmath31 ( an oil spot on the plane ? ) causes the invalidity of the second relation of ( [ condizioni_cm_es1 ] ) , while the first relation still holds .
of course , even if we chose the assumption of step 4 of the procedure as a fixed point of our argument , we can not suppose that the system continues to roll in @xmath30 ( and begins to slide in @xmath31 ) since the beginning of sliding in @xmath31 can affect the pure rolling behavior of the system in the point @xmath30 .
we must perform a new analysis of the behavior of the system , possibly supposing the system rolling in @xmath30 and sliding in @xmath31 , we must determine ( if possible ) the new equations of motion of the system ( with the additional difficulties of different friction laws in the point @xmath30 and @xmath31 and possibly increased number of degrees of freedom ) , the motion of the system , the new ( parametric ) system of reactions acting on the system and then we can test the coulomb condition in the point @xmath30 . | we illustrate a theoretical procedure determining necessary conditions for which simultaneous pure rolling kinetic constraints acting on a mechanical system can be fulfilled .
we also analyze the sufficiency of these conditions by generalizing to this case a well known and usually accepted assumption on the behavior of pure rolling constraint .
we present in detail the application of the procedure to some significative mechanical systems .
0.5truecm * pacs : * 45.40.-f , 45.50.-j * 2000 mathematical subject classification : * 70e18 70f25 70e60 * keywords : * rolling constraint linear and angular momenta equations |
effective field theory ( eft ) is a powerful framework based on exploiting symmetries and controlled expansions for problems with a natural separation of energy or distance scales .
efts are particularly important in qcd and nuclear physics .
an effective theory of qcd , ideally suited to jet applications , is soft - collinear effective theory ( scet ) @xcite .
recently , first steps were taken to extend scet and describe jet evolution in strongly - interacting matter @xcite by introducing a transverse momentum exchange glauber gluon mode between the collinear partons and the medium quasi - particles .
the newly constructed theory , called soft - collinear effective theory with glauber gluons ( @xmath2 ) , was used to derive all @xmath7 @xmath8 medium - induced splitting kernels and discuss higher order @xmath9 corrections to the medium - modified jet substructure @xcite .
these medium - induced splitting kernel are the universal key ingredients in the evaluation of a wide variety of hadronic and jet observables in nucleus - nucleus collisions .
the theoretical advances reported here allowed us for the first time to go beyond the traditional energy loss approximation to parton propagation in matter and to unify the treatment of vacuum and medium - induced parton showers @xcite .
they provide quantitative control over the uncertainties associated with the implementation of the in - medium modification of hadron production cross sections and help accurately constrain the coupling between the jet and the medium .
elementary parton branching is the essential steps in the formation of a parton shower .
it was demonstrated that in the ambiance of dense qcd matter full splitting functions are equal to the sum of the vacuum ones and the corresponding in - medium contribution : @xmath10 where @xmath11 is the momentum fraction carried by the daughter parton , @xmath12 is the transverse momentum relative to the parent parton , and @xmath13 collectively describes the medium properties .
the splitting functions factorize from the hard scattering process and are gauge invariant .
they govern the evolution of parton distribution functions and fragmentation functions ( ffs ) in the medium . here , we are interested in final - state interactions and the generalized dglap evolution equations for ffs in the medium are written down as follows @xcite @xmath14\ , , \label{eq : map10 } \\
\frac{d d_{h / g}(z , q)}{d \ln q}&=&\frac{\alpha_s(q)}{\pi}\int_{z}^1 \frac{d z'}{z'}\big[p^{\rm med}_{g\rightarrow gg}(z ' , q ; \beta)d_{h / g}\left(\frac{z}{z'},q\right ) % \nonumber \\ & & + p^{\rm med}_{g\rightarrow q\bar q}(z ' , q ; \beta)\sum_q d_{h
/ q}\left(\frac{z}{z'},q\right)\big]\ , .
\label{eq : map30}\end{aligned}\ ] ] these evolution equations also encode the effect of parton energy loss from multiple soft gluon emissions . in the @xmath15 limit , the solution of eqs .
( [ eq : map10 ] ) , ( [ eq : map30 ] ) is related to the traditional parton energy loss picture @xcite .
@xmath16 at rhic with @xmath17 are compared to the phenix suppression measurements in central au+au collisions .
right panel : nuclear modification factor @xmath16 for charged hadrons is calculated in central pb+pb collisions at the lhc @xmath4 atev , and compared with the atlas experimental data .
, title="fig:",width=283 ] @xmath16 at rhic with @xmath17 are compared to the phenix suppression measurements in central au+au collisions .
right panel : nuclear modification factor @xmath16 for charged hadrons is calculated in central pb+pb collisions at the lhc @xmath4 atev , and compared with the atlas experimental data .
, title="fig:",width=264 ] the medium - evolved ffs can be implemented in phenomenological calculations of hadron suppression in heavy ion collisions .
comparison of the four different evaluations of the nuclear modification @xmath16 for rhic @xmath0 agev in central au+au reactions is shown in the left panel of fig .
[ fig : rhicraa ] .
the calculations provide adequate description of the attenuation of the inclusive @xmath18 cross section measured by the phenix experiment @xcite .
the difference between the full solution to the dglap evolution equations , semi - analytic solutions in the soft gluon limit , and the traditional energy loss approach are very small and corresponds to @xmath19 uncertainty in the determination of the coupling between the jet and the medium .
results for the suppression of inclusive charged hadron production at @xmath4 atev in central pb+pb collisions at the lhc are shown in the right panel of fig .
[ fig : rhicraa ] and atlas data is included for comparison .
calculations based on solutions of the evolution equations with in - medium splitting kernels give good description of the centrality and transverse momentum @xmath20 dependence of charged hadron production at the lhc , while for neutral pions measurements to higher @xmath20 will help better assess if any discrepancy between theory and measurements exists @xcite .
at leading order , the jet energy function associated with parton @xmath21 with the collinear momentum @xmath22 splitting into @xmath23 and @xmath24 can then be written as follows , @xmath25 where @xmath26 are the collinear parton splitting functions . here , @xmath27 is the measurement function associated with the jet energy function . since the splitting functions are directs sums of their vacuum and medium - induced components eq .
( [ gensum ] ) , we find @xmath28j^{i}_{\omega , e_r}(\mu ) \ , , \ ] ] where @xmath29 with @xmath30 and @xmath31 the casimir operators of the fundamental and adjoint representations in qcd , and @xmath32 and @xmath33 are the cusp and non - cusp anomalous dimensions @xcite .
furthermore , the in - medium splitting functions can be used to generalize the concept of energy loss for jets with reconstruction parameter @xmath3 beyond the soft gluon approximation .
more specifically , the fractional medium - induced energy loss of quark and gluon jets is @xmath34\int_{\omega x(1-x)\tan\frac{r}{2}}^{\omega x(1-x)\tan\frac{r_0}{2}}dk_\perp \frac{1}{2 } \big [ { \cal p}^{med}_{q\rightarrow qg}(x , k_\perp ) + { \cal p}^{med}_{q\rightarrow gq}(x , k_\perp ) \big ] \ ; , \\ \epsilon_g & = & \frac{2}{\omega}\big[\int_0^{\frac{1}{2 } } dx k^0 + \int_{\frac{1}{2}}^1 dx ( p^0-k^0 ) \big]\int_{\omega x(1-x)\tan\frac{r}{2}}^{\omega x(1-x)\tan\frac{r_0}{2}}dk_\perp \frac{1}{2 } \big [ { \cal p}^{med}_{g\rightarrow gg}(x , k_\perp)+ \sum_{q,\bar{q}}{\cal p}^{med}_{g\rightarrow q\bar q}(x , k_\perp)\big].\end{aligned}\ ] ] note that in the splitting @xmath35 the final - state partons are identical particles .
here @xmath3 is the angular parameter used in the jet reconstruction , and @xmath36 is of @xmath37 in qcd which sets the region of the use of collinear parton splitting functions . as a function of the jet transverse momentum to experimental data in pb+pb collisions at @xmath4 atev
are compared to alice , atlas and cms data .
right panel : modification of differential jet shapes of inclusive jets at the lhc .
the modification is presented as the ratio of the jet shapes @xmath38 in pb+pb and p+p collisions .
the blue band corresponds to the calculations including only the cnm effects , the red band adds the jet - medium interaction but with the jet - by - jet shape modification turned off , and the green band correspond to the full calculation .
data is from cms with @xmath39 .
, title="fig:",width=279 ] as a function of the jet transverse momentum to experimental data in pb+pb collisions at @xmath4 atev are compared to alice , atlas and cms data .
right panel : modification of differential jet shapes of inclusive jets at the lhc .
the modification is presented as the ratio of the jet shapes @xmath38 in pb+pb and p+p collisions .
the blue band corresponds to the calculations including only the cnm effects , the red band adds the jet - medium interaction but with the jet - by - jet shape modification turned off , and the green band correspond to the full calculation .
data is from cms with @xmath39 .
, title="fig:",width=279 ] the results for the medium - modified jet energy functions and reconstructed jet energy loss can be implemented in phenomenology .
the left panel of fig .
[ fig : jetmod ] shows the suppression of inclusive jet production in central pb+pb collisions at the lhc .
the bands correspond to couplings between the jet and the medium @xmath40 and no cold nuclear matter ( cnm ) energy loss , small cnm e - loss , and moderate cnm e - loss @xcite .
experimental data is from the alice , atlas and cms collaborations @xcite . in the right panel of fig .
[ fig : jetmod ] comparison of the calculated jet shape modification ( green band ) to cms results at @xmath4 atev is shown .
the contribution of various initial - state and final - state effects is also presented .
in central and mid - peripheral pb+pb collisions at @xmath41 atev .
right panel : related modification of the differential jet shapes for inclusive jets ( green band ) and photon - tagged jets ( blue band ) , with @xmath42 in central pb+pb collisions . , title="fig:",width=234 ] in central and mid - peripheral pb+pb collisions at @xmath41 atev .
right panel : related modification of the differential jet shapes for inclusive jets ( green band ) and photon - tagged jets ( blue band ) , with @xmath42 in central pb+pb collisions . , title="fig:",width=279 ] in summary , we presented results for the suppression of inclusive hadron production in au+au reactions at rhic and pb+pb reactions at the lhc based upon qcd factorization and dglap evolution with @xmath2-based medium - induced splitting kernels .
this method allowed us to unify the treatment of vacuum and medium - induced parton showers .
in the soft gluon bremsstrahlung limit , we demonstrated the connection between this new approach and the traditional energy loss - based jet quenching phenomenology . with an emphasis on a consistent theoretical descriptions of hadron and jet observables in heavy ion collisions
, we further related the in - medium modification of parton showers to the attenuation of reconstructed jet cross sections and the modification of the jet shapes in heavy ion collisions .
while good description of the majority of current experimental measurements is achieved , this theoretical framework can be further tested by the upcoming lhc run ii measurements .
we present predictions for charged hadron and neutral pion @xmath16 , the jet shape modification and the cross section suppression for inclusive jets and photon - tagged jets see fig .
[ fig : pred ] .
we fond that the cross section suppression at high @xmath20 can provide information about the cold nuclear matter effects .
since photon - tagged jets are predominately quark - initiated , the cross section is expected to be less suppressed compared to inclusive jets . on the other hand , the broadening of the photon - tagged jet is more apparent . c. w. bauer , s. fleming and m. e. luke , phys .
d * 63 * ( 2000 ) 014006 ; c. w. bauer , s. fleming , d. pirjol and i. w. stewart , phys .
d * 63 * , 114020 ( 2001 ) .
a. idilbi and a. majumder , phys .
d * 80 * ( 2009 ) 054022 ; f. deramo , h. liu and k. rajagopal , phys .
d * 84 * ( 2011 ) 065015 ; g. ovanesyan and i. vitev , jhep * 1106 * ( 2011 ) 080 .
z. b. kang , r. lashof - regas , g. ovanesyan , p. saad and i. vitev , phys .
* 114 * ( 2015 ) 9 , 092002 ; y. t. chien , a. emerman , z. b. kang , g. ovanesyan and i. vitev , arxiv:1509.02936 [ hep - ph ] ; y. t. chien and i. vitev , arxiv:1509.07257 [ hep - ph ] .
a. adare _ et al .
_ [ phenix collaboration ] , phys .
* 101 * ( 2008 ) 232301 . b. abelev
_ et al . _
[ alice collaboration ] , phys .
b * 720 * ( 2013 ) 52 ; b. b. abelev _ et al .
_ [ alice collaboration ] , eur .
j. c * 74 * ( 2014 ) 10 , 3108 ; g. aad _ et al .
_ [ atlas collaboration ] , jhep * 1509 * ( 2015 ) 050 ; s. chatrchyan _ et al . _
[ cms collaboration ] , eur .
j. c * 72 * ( 2012 ) 1945 .
i. vitev , phys .
c * 75 * ( 2007 ) 064906 ; z. b. kang , i. vitev and h. xing , phys . rev .
c * 92 * ( 2015 ) 5 , 054911 .
b. abelev _ et al . _
[ alice collaboration ] , jhep * 1403 * ( 2014 ) 013 ; g. aad _ et al .
_ [ atlas collaboration ] , phys .
b * 719 * ( 2013 ) 220 ; g. aad _ et al .
_ [ atlas collaboration ] , phys .
* 114 * ( 2015 ) 7 , 072302 ; y - j .
lee [ cms collaboration ] , private communication . | in the framework of soft - collinear effective theory with glauber gluons , results and predictions for inclusive hadron suppression , based upon in - medium parton shower evolution , are presented for au+au and pb+pb collisions at rhic and lhc energies @xmath0 agev and @xmath1 atev , respectively . the @xmath2 medium - induced splitting kernels are further implemented to evaluate the attenuation of reconstructed jet cross in such reactions and to examine their centrality and radius @xmath3 dependence . building upon a previously developed method to systematically resum the jet shape at next - to - leading logarithmic accuracy , a quantitative understanding of the jet shape modification measurement in pb+pb collisions at @xmath4 atev at the lhc can be achieved .
predictions for photon - tagged jet cross sections and shapes , that can shed light on the parton flavor dependence of in - medium parton shower modification , are also given .
% @xmath5 , inclusive hadron suppression , qcd evolution , reconstructed jet quenching , jet shapes , @xmath6-tagged jets |
the dimension of the multiwire chambers deployed in modern high energy physics experiments is usually large conforming to the scale of experimental setup .
the electrostatic instability in such chambers may be crucial when the amplitude of the oscillation caused by the action of electrostatic force alone or combined with the gravity becomes comparable to the electrode spacings .
the study of the wire deflection in such a geometry is usually a complex affair since an interplay between several physical forces determines the wire stability .
the approximation of constant or linear dependence of the force on the wire deflection is not adequate to solve for the differential equation governing the wire dynamics because all the wires in the chamber move in a collective way influencing each other giving rise to a nonlinear effect .
since the exact solutions for the differential equation involving the nonlinear force are no longer known , it has to be solved numerically . of various methods of estimating the electrostatic sag from the differential equation ,
only the linear and iterative methods have been attempted in several geometries @xcite . in these works ,
the electrostatic force has been estimated from the 2d field calculation @xcite which differs significantly from 3d solutions .
owing to the 2d nature of the problem , the sag is normally overestimated due to the fact that the whole length of the wire is considered to be at maximum sag . in this work ,
an accurate 3d computation of electrostatic field has been carried out through the use of a nearly exact boundary element method ( nebem ) @xcite which has yielded precise force estimation . in order to reduce complexity ,
only the normal component of the field has been considered in the calculation .
the deflection of each segment has been assumed to be very small in comparison to its length .
the calculation has been carried out for a geometry similar to that of rich detector in alice @xcite .
the anode plane consists of gold - tungsten wires with @xmath0 m diameter with pitch @xmath1 mm .
the upper cathode plane is made of copper - berrylium wires with diameter @xmath2 m and pitch @xmath3 mm while the lower one is a uniform conducting plate .
the separation of upper and lower cathodes from the anode are respectively @xmath4 mm and @xmath5 mm and length of the detector in z - direction is @xmath6 cm .
the anode plane is supplied with high voltage w.r.t . the cathode planes .
the second order differential equation in an equilibrium state of the wire can be written as @xmath7 where @xmath8 , @xmath9 are the electrostatic and gravitational forces per unit length while @xmath10 the stringing tension of the wire . using three point finite difference formula , it can be rewritten as @xmath11.(\delta z)^2\ ] ] where @xmath12 , @xmath13 and @xmath14 represent the deflections of respective segments . the electrostatic force on the @xmath15-th segment has been computed using nebem solver for the given 3d geometry .
the required sag due to the action of either of the electrostatic and gravitational forces or combined may be obtained from this equation .
thus the set of equations for the segments on a wire can be represented as @xmath16 where @xmath17 is the tridiagonal coefficient matrix whose inverse has been calculated following standard numerical receipe . in the present work ,
five anode wires have been considered with discretization of @xmath18 linear segments while that of the cathode plate has been @xmath19 .
it should be noted that no plates on the sides of the chamber have been taken into account .
the calculation procedure has been validated by calculating wire sag due to gravitational force and comparing with the analytic solution for gravitational force only as @xmath20 where @xmath21 and @xmath22 , @xmath23 , @xmath24 are the length , radius and density of the wire respectively .
the results has been illustrtaed in fig.[fig : gravsagandcath ] which has demonstrated the validity of the method . gravitational sag of central anode and cathode wires , scaledwidth=45.0% ]
the normal electric field components acting on the anode and cathode wire segments for anode voltage of @xmath25v have been plotted in fig.[fig : normalef ] .
the field component on each segment has been calculated from the vectorial addition of field components at four radial locations on the segment periphery .
the wire sag at the centre due to electrostatic force following the solution of tridiagonal matrix equation [ eqn.[eqn : mateq ] ] has been shown as a function of anode voltage in fig.[fig : wiresag ] for anode and cathode wires separately .
it is evident from the result that the sag in the anode wire changes more rapidly than the cathode wires .
r0.5 the central wire in the anode plane has been found to undergo more deflection in comparison to the edge wires .
the calculation of @xcite for wire sags in this chamber has reported less deflection in comparison to our result . in @xcite ,
an additional restoring electrostatic force has been considered to be operational when the wire gets deflected which in turn has helped to reduce the wire sag . in our calculation , no such dynamic consideration of the electrostatic force with the wire deflection has been incorporated . to reproduce the actual wire sags ,
an iterative process can be carried out each time calculating the electrostatic force due to new position of the deflected wire .
using the nebem solver , the electrostatic field could be accurately calculated for the three dimensional geometry of multiwire rich chamber .
an fdm approach to compute the wire sag has been developed and validated for the case of gravitational sag calculation . in the present calculation ,
no restoring effect of electrostatic force has been considered unlike the earlier work which has led to larger sag estimates .
the restoring force aspect will be implemented in future by iterative technique to estimate a realistic wire sag in this chamber . | a numerical method of determining the wire sag in a multiwire proportional chamber used in rich @xcite by solving the second order differential equation which governs the wire stability has been presented .
the three point finite difference method ( fdm ) has generated a tridiagonal matrix equation relating the deflection of wire segments to the force acting on it .
the precise estimates of electrostatic force has been obtained from accurate field computation using a nearly exact boundary element method ( nebem ) solver @xcite . |
g. stokes studied the polarization properties of a quasi - monochromatic plane wave of light in an arbitrary polarization state by introducing four quantities , known since then as the stokes parameters @xcite .
the stokes parameters are a set of four quantities which provide intuitive and practical tools to describe the polarization of light @xcite .
the stokes parameters give a direct relation between the light ( photons ) polarization and the polarization of elementary particles @xcite .
this fact was widely exploited to study many features of radiation of particles and to the scattering problems @xcite .
the stokes parameters are formulated in terms of the observables of the electromagnetic field , namely , the amplitudes and the relative phase difference between the orthogonal components of the field @xcite .
in fact , the density matrix @xcite and the coherence matrix @xcite for the case of electromagnetic radiation result to be the same @xcite , and are written in terms of these observables . the standard procedure to describe the polarization of an electromagnetic wave is to set the propagation direction along the @xmath1-axis , and the two components of the polarization field on the @xmath6 and @xmath7directions .
however , when the direction of arrival from the source is unknown a priori , the three - dimensional coherence matrix must be used to obtain a complete polarization characterization @xcite . in ref .
@xcite , jauch and rohrlich introduced the stokes parameters in the quantum regime , which are called stokes operators .
it is at the quantum domain where we can see that a symmetry group structure is related to the stokes operators . when the direction of propagation of light is known , the symmetry is the @xmath8 group @xcite
however , when the direction of propagation is unknown , the symmetry group is @xmath9 @xcite . also , other generalizations of stokes operators have been reported @xcite . in this work
we give a new approach to the theory of light polarization . for simplicity , we study the case of a monochromatic plane electromagnetic wave which propagates along the @xmath1-axis .
our study is based on a set of quantum operators , constructed from two independent bosons , being three of them the @xmath0 lie algebra generators , and the other one , the casimir operator of this algebra .
this work is organized as follows . in section 2
, we deduce the @xmath0 lie algebra generators by the jordan - schwinger map . in section 3 , by taking the expectation value of the algebra generators in a two - mode coherent state , we obtain their classical limit . in section 4 , we define our stokes parameters ( we refer to them stokes - like parameters ) and show that the light polarization ellipse can be written in terms of them . in section 5 , the stokes - like parameters are written in terms of two parameters and it is shown that they define a one - sheet ( poincar hyperboloid ) of a two - sheet hyperboloid .
in what follows we will use @xmath10 , where @xmath11 is the mass of each one - dimensional harmonic oscillator and @xmath12 is the frequency of either the electromagnetic wave or the harmonic oscillators .
we define the operators @xmath13 with @xmath14 the operators @xmath15 and @xmath16 the left and right annihilation operators of the two - dimensional harmonic oscillator , with the non vanishing commutators @xmath17=[a_2,a_2^{\dag}]=1 $ ] .
the matrices @xmath18 are defined as follows : @xmath19 , @xmath20 , @xmath21 and @xmath22 , where @xmath23 are the usual pauli matrices @xcite .
+ explicitly , the operators @xmath24 , @xmath25 , @xmath26 and @xmath27 are given by @xmath28 the operator @xmath24 can be rewritten as @xmath29 , being @xmath30 the @xmath31component of the angular momentum of the two - dimensional harmonic oscillator , whose hamiltonian is given by @xmath32 .
therefore , the operator @xmath27 is essentially the energy of the two - dimensional harmonic oscillator .
it can be shown that @xmath33=0 , \hspace{2ex}\ ] ] also , by a straightforward calculation , we show that the commutation relations of the operators @xmath34 are @xmath35=-i \mathcal{k}_3,\hspace{3ex } \left[\mathcal{k}_2,\mathcal{k}_3\right]=i \mathcal{k}_1 , \hspace{3ex } \left[\mathcal{k}_3,\mathcal{k}_1\right]=i \mathcal{k}_3 . \label{comu}\ ] ] therefore , these operators close the @xmath0 lie algebra . the casimir operator @xmath36 for this algebra results to be @xmath37 .
hence , we have obtained the generators of the @xmath0 lie algebra , equation ( [ jsm ] ) , by the so called jordan - schwinger map @xcite .
in ref . @xcite the classical states of the two - dimensional harmonic oscillator were calculated . on the other hand ,
the stokes parameters were obtained by evaluating the @xmath38 lie algebra generators in a two - mode coherent state @xcite .
the same idea was used to derive one of the stokes parameters generalizations by evaluating the @xmath39 lie algebra generators in a three - mode coherent states @xcite . in this paper
we take advantage of these facts .
thus , we evaluate the @xmath0 lie algebra generators @xmath40 in a two - mode coherent state to obtain their classical limit . the two - mode coherent states are well known to be given by @xmath41}\sum_{n_1,n_2=0}^\infty{\alpha_1^{n_1}\alpha_2^{n_2 } \over \sqrt{n_1!n_2!}}|n_1,n_2\rangle,\label{tm}\ ] ] which are eigenstates of the annihilation operators @xmath15 and @xmath16 @xmath42 with this procedure , we obtain @xmath43 where @xmath44 are the classical oscillations @xmath45 $ ] with amplitudes @xmath46 and phases @xmath47 .
this leads to @xmath48 where @xmath49 .
thus , the classical limit of the lie algebra generators in a time - dependent two - mode coherent state is time dependent .
this is because the @xmath0 lie algebra generators do not commute with the hamiltonian of the two - dimensional harmonic oscillator .
it is well known that the standard stokes parameters obtained as the classical limit of the @xmath38 generators are time - independent @xcite .
this is due to the @xmath38 lie algebra generators commute with the two - dimensional harmonic oscillator hamiltonian .
we set an electromagnetic wave with arbitrary polarization , which propagates along the @xmath1-axis , given by @xmath50 where the complex amplitudes are defined by @xmath51}$ ] .
it can be written in the form @xmath52 where @xmath53 and @xmath54 from these equations is easy to obtain the nonparametric quadratic equation @xmath55 by setting the definitions @xmath56 , @xmath57 and @xmath58 , it is shown that @xmath59 .
thus , when @xmath60 , equation ( [ trayec ] ) represents an ellipse centered at the origin . equations ( [ campo1 ] ) and ( [ campo2 ] ) allows to write equation ( [ trayec ] ) as @xmath61 this is the polarization ellipse described by the plane wave electromagnetic field . using equations ( [ funda1 ] ) and ( [ funda2 ] ) , and the definition of the phase difference @xmath62 , this equation can be rewritten as @xmath63 in this way we have written the polarization ellipse in terms of the classical values of the @xmath0 lie algebra generators .
+ we define the classical stokes - like parameters @xmath64 as follows @xmath65 thus , the ellipse polarization can be expressed as @xmath66 we note that the polarization ellipse , equation ( [ elipse ] ) , depends on the phase difference @xmath67 , whereas the classical limit of the @xmath0 generators do not .
this fact has forced us to define the stokes - like parameters as those of equations ( [ stokes1 ] ) and ( [ stokes2 ] ) , which depends on the phase difference and are time - independent .
the procedure to map the standard stokes parameters to the poincar sphere is explained in full detail in sec .
1.4.2 of ref . @xcite . in the present section ,
we translate the born and wolf ideas to our stokes - like parameters .
we shall show that they are mapped to a one - sheet of a two - sheet hyperbolic paraboloid .
we perform an hyperbolic rotation to the fields @xmath68 and @xmath69 by an angle @xmath70 , in such a way that the fields @xmath71 and @xmath72 are obtained on the principal axis of the ellipse .
thus , @xmath73 equation ( [ campo ] ) can be written as @xmath74 where @xmath75 , and @xmath76 , @xmath77 .
+ on the other hand , the oscillations on the principal axis of the ellipse are given by @xmath78 being @xmath79 a constant phase .
the signs @xmath80 are for right and left polarization on the coordinates system @xmath81 . from the substitution of equations ( [ oscilaciones ] ) and ( [ oscilaxi ] ) into equation ( [ rot1 ] )
, we deduce the equalities @xmath82 similarly , by substituting equations ( [ oscilaciones ] ) and ( [ oscilaeta ] ) into equation ( [ rot2 ] ) , we deduce @xmath83 dividing ( [ ocho ] ) by ( [ seis ] ) and ( [ nueve ] ) by ( [ siete ] ) , we obtain @xmath84 the second equality and some elementary hyperbolic identities lead to @xmath85 the definition @xmath86 and the identity for @xmath87 permit to write this equation as @xmath88 on squaring and adding ( [ seis ] ) and ( [ siete ] ) , we obtain @xmath89 also , on squaring and adding ( [ ocho ] ) and ( [ nueve ] ) : @xmath90 the subtracting of equation ( [ bcuad ] ) from equation ( [ acuad ] ) allows to obtain @xmath91 if we multiply ( [ seis ] ) by ( [ ocho ] ) , and ( [ siete ] ) by ( [ nueve ] ) and adding the respective results , we get @xmath92 from the last two equalities it is implied that @xmath93 if we introduce the definition @xmath94 , this equation result to be @xmath95 now , the definition @xmath96 above , leads to @xmath97 equation ( [ funda4 ] ) can be written in terms of equations ( [ funda1 ] ) , ( [ funda2 ] ) and ( [ oscilaciones ] ) as @xmath98 where for the last equality , we have used equation ( [ stokes1 ] ) .
this equation leads to @xmath99 or depending on the sign of the @xmath100 : @xmath101 similarly , from equations ( [ funda3 ] ) , ( [ funda1 ] ) , ( [ funda2 ] ) , and ( [ stokes2 ] ) , we can show that @xmath102 on the other hand , from equations ( [ funda1 ] ) and ( [ funda2 ] ) it is immediate to show that @xmath103 where for the last equality we have used equations ( [ stokes1 ] ) and ( [ stokes2 ] ) .
equations ( [ tild1 ] ) , ( [ tild2 ] ) and ( [ forc ] ) lead to @xmath104 however , because of equation ( [ stokes2 ] ) , @xmath105 , thus @xmath106 substituting this equation into equation ( [ tild2 ] ) we have @xmath107 equations ( [ tild1 ] ) , ( [ tild3 ] ) and ( [ tild4 ] ) are the stokes - like parameters written in terms of @xmath108 and @xmath70 .
we have plotted the vector function @xmath109 with the different setting of signs : a ) ( + , + , + ) , b ) ( + , ,+ ) , c ) ( ,+,+ ) and d ) ( ,,+ ) .
it must be emphasized that for each of these setting , the polarization surface generated for the stokes - like parameters results to be as that shown in fig . 1 , been @xmath110 the distance from the origin of coordinates to the minimum of the hyperboloid sheet .
hence , because of the plus sign in equation ( [ tild3 ] ) , the stokes - like parameters generate a one - sheet ( poincar hyperboloid ) of a two - sheet hyperboloid , instead an sphere . fig .
1 shows how the poincar hyperboloid varies as a function of @xmath110 . + .
the greater one is for @xmath111 and the little one is for @xmath112 . ]
[ f1 ] now , we are interested in investigating the regions on the poincar hyperboloid in which occurs the different polarizations : i)linear polarization @xmath113 , ii)circular polarization @xmath114 and iii ) elliptical polarization @xmath115 .
+ i ) for linear polarization @xmath113 , @xmath116 , @xmath117 @xcite , equations ( [ stokes1 ] ) and ( [ stokes2 ] ) imply that @xmath118 and @xmath119 to a real quantity .
also , for @xmath116 , @xmath120 a linear polarization holds . for this case ,
@xmath118 and @xmath119 to a real quantity .
in this way , we have showed that each point on the curve @xmath113 of fig .
2 represents to a linear polarization .
+ ii ) the circular polarization @xmath114 , occurs when @xmath121 , ( @xmath80 represents right ( left ) circular polarization , respectively ) , @xmath122 and @xmath123 . in order to have a graphical interpretation to this case
, we consider a limit process @xmath124 and @xmath125 . under this assumption : a ) the minimum of the poincar hyperboloid approaches to the origin of coordinates , as shown in fig . 1 , and
b ) with @xmath121 , equations ( [ funda1 ] ) , ( [ funda2 ] ) , ( [ stokes1 ] ) and ( [ stokes2 ] ) implies that @xmath126 and @xmath127 .
in fact , for right circular polarization , @xmath128 , @xmath129 and @xmath127 , and for left circular polarization , @xmath130 , @xmath131 and @xmath127 . hence , each point of the curve @xmath114 on the poincar hyperboloid represents a circular polarization ( see fig .
+ iii ) it is well known that right elliptical polarization @xmath132 holds for @xmath133 , and left elliptical polarization @xmath134 holds for @xmath135 @xcite . since for @xmath132 , @xmath136 , from equations ( [ stokes1 ] ) and ( [ stokes2 ] ) we deduce that @xmath137 whereas @xmath138 can takes positive and negative values . similarly , for @xmath134 , @xmath139 and @xmath138
can takes both positive and negative values . in fig
2 we have represented these two regions . .
we have set @xmath140 , @xmath141 and @xmath142 .
the curve @xmath113 represents the points where linear polarization occurs , and in each point on the curve @xmath114 there exists a circular polarization .
each point on the region @xmath132 represents a right elliptical polarization , and the points on the region @xmath134 are those on which left elliptical polarization holds .
] [ f2 ]
we have given an alternative definition of the stokes parameters ( stokes - like parameters ) and showed that the polarization ellipse can be written in terms of these parameters .
based on the procedure to map the standard stokes parameters to the poincar sphere @xcite , we achieve to map our stokes - like parameters to a one - sheet ( poincar hyperboloid ) of a two - sheet hyperboloid .
we note that in the equations ( [ stokes1 ] ) and ( [ stokes2 ] ) , the definitions of @xmath143 and @xmath138 can be interchanged without any essential change in the results . finally , we emphasize that the case we have treated in this paper is when the direction of arrival of the electromagnetic wave is known a priori .
our results rest on the lie algebra @xmath144 .
when the direction of arrival of the electromagnetic wave is unknown a priori , a generalization of the results of the present paper could be related with the generators of the @xmath145 lie algebra .
this work was partially supported by : cofaa - ipn , edi - ipn , edd - ipn and cgpi - ipn project numbers 20161727 and 20160108 . 99 stokes g. g. , trans .
cambridge philos .
soc . , * 9*,(1852 ) 399 ( reprinted in stokes g. g. , mathematical and physical papers , johnson reprint corporation , n. y. and london , 1966 ) .
shurcliff w. a. and ballard s. s. , polarized light , van nostrand company , princeton , n.j . , 1964 .
mcmaster w. h. , am . j. phys . * 22 * , ( 1954 ) 351 .
mcmaster w. h. , rev .
* 33 * , ( 1961 ) 8 .
collett e. , am . j. phys .
* 36 * , ( 1968 ) 713 .
wolf e. , nuovo cimento * 12 * , ( 1954 ) 884 .
wolf e. , nuovo cimento , * 13 * , ( 1959 ) 1165 .
blaskal s. and kim y. s. , arxiv : quant - ph/0501050 .
roman p. , nuovo cimento * 13 * , ( 1959 ) 974 .
ramachandran g. , murthy m. v. n. and mallesh k. s. , pramana * 15 * , ( 1980 ) 357 .
carozzi t. , karlsson r. and bergman j. , phys .
e * 65 * , ( 2000 ) 2024 .
mota r. d. , xicotncatl m. a. and granados v. d. , j. phys a : math.and gen . *
37 * , ( 2004 ) 2835 .
jauch j. m. and rohrlich f. , the theory of photons and electrons , springer - verlag , berlin , 1976 .
mota r. d. , xicotncaltl m. a. and granados v. d. , can . j. phys . * 82 * , ( 2004 ) 767 .
abouraddy a. f. , sergienko a. v. , saleh b. e. a. and teich m. c. , opt . commun . * 201 * , ( 2002 ) 93 .
jaeger g. , teodorescu - frumosu m. , sergienko a. b. , saleh b. e. a. and teich m. c. , phys .
rev . a * 67 * , ( 2003 ) 032307 .
biedenharn l. c. and louck j. d. , angular momentum in quantum physics , addison - wesley publishing company , massachusetts , 1981 .
cohen - tannoudji c. , diu b. and lale f. , quantum mechanics , john wiley @xmath146 sons , paris , 1977 .
hecht e. , optics , addison wesley , san francisco , 2002 .
born m. and wolf e. , principles of optics , pergamon press , oxford , 1975 . | we introduce an alternative approach to the polarization theory of light .
this is based on a set of quantum operators , constructed from two independent bosons , being three of them the @xmath0 lie algebra generators , and the other one , the casimir operator of this algebra . by taking the expectation value of these generators in a two - mode coherent state
, their classical limit is obtained .
we use these classical quantities to define the new stokes - like parameters .
we show that the light polarization ellipse can be written in terms of the stokes - like parameters .
also , we write these parameters in terms of other two quantities , and show that they define a one - sheet ( poincar hyperboloid ) of a two - sheet hyperboloid . our study is restricted to the case of a monochromatic plane electromagnetic wave which propagates along the @xmath1 axis .
@xmath2 escuela superior de ingeniera mecnica y elctrica , unidad culhuacn , instituto politcnico nacional , av .
santa ana no .
1000 , col .
san francisco culhuacn , delegacin coyoacn , c.p .
04430 , mxico d. f. , mexico .
+ @xmath3 escuela superior de cmputo , instituto politcnico nacional , av . juan de dios batiz , esq .
av .
miguel othon de mendizbal , col .
lindavista , del .
gustavo a. madero , c.p .
07738 , mxico d. f. , mexico .
+ @xmath4 escuela superior de fsica y matemticas , instituto politcnico nacional , ed .
9 , unidad profesional adolfo lpez mateos , c.p .
07738 , mxico d. f. , mexico .
+ keywords : @xmath5 lie algebra , light polarization , poincar hyperboloid |
solitary - waves may be found in solutions to nonlinear wave equations where the nonlinearity counteracts the dispersion of a wave - packet such that it retains its form as it propagates .
solitons are solitary - waves that emerge unscathed from collisions with each other , up to shifts in position and phase ; this behaviour is reminiscent of particle behaviour , motivating the particle - like name soliton .
this distinction is an important one , although in practice the names soliton and solitary - wave are commonly interchanged .
`` classic '' solitons , in this sense , are to be found in integrable nonlinear wave - equations , such as the korteweg - de vries equation , the sine - gordon equation , and the one - dimensional nonlinear schrdinger equation .
the solitons ability to re - emerge after collisions is due to the fact that their dynamics are strongly constrained by conservation laws associated with the wave - equations integrability @xcite .
solitons and solitary - waves are topics of keen interest in the atomic bose - einstein condensate ( bec ) community .
this is because low - temperature bec dynamics are frequently described to a good approximation by the gross - pitaevskii equation ( gpe ) @xcite , a 3d nonlinear wave equation . for regimes where the atoms are confined in the radial direction by a tight trapping potential
, the 3d gpe reduces approximately to a 1d equation ( the so - called 1d gpe ) .
the homogeneous 1d gpe is simply the 1d nonlinear schrdinger equation , which can be solved by the inverse scattering transform , and yields bright soliton solutions when the nonlinearity is attractive @xcite . at sufficiently low temperatures
the interatomic scattering can be largely described through a single parameter , the @xmath0-wave scattering length . in this context
, an attractive nonlinearity arises from a negative @xmath0-wave scattering length , which may be intrinsic , or which may be induced by exploiting a feshbach resonance to tune the inter - atomic interactions @xcite . as well as describing bec under tight transverse confinement
, the 1d nonlinear schrdinger equation is also used to describe nonlinear optical systems @xcite .
these systems provide a useful analogue of bec under tight transverse confinement , and we will frequently refer to work on nonlinear optics in this paper . experiments involving becs composed of attractively interacting atoms have been performed in 1d geometries , resulting in the observation of single @xcite and multiple bright solitary - waves @xcite . in the experiments with multiple solitary - waves ,
the bec was trapped in the axial direction by a ( relatively weak ) harmonic confining potential in addition to the radial confinement .
the addition of an axial harmonic potential acts to break the integrability of the 1d gpe , meaning that we no longer have exact soliton solutions . in the experiment by strecker _
@xcite , classic soliton - like behaviour ( where the solitary - waves collide and reform up to shifts in phase and position ) was not observed , but rather , trains of solitary - waves which are continuously repelled by each other .
the dynamics of solitary - wave trains both in bec and nonlinear optics have been the topic of extensive modeling using a variational method @xcite , numerical simulations @xcite , a toda lattice approach , @xcite , a particle model @xcite ( quite distinct to that presented in this paper ) , analysis using the inverse - scattering transform @xcite and by using a perturbation approach @xcite .
these treatments model regimes where the solitary - waves are never well separated , where it has been found that the solitary - waves do not collide with each other and re - form , but interact with each other by attractive and repulsive forces , depending on their relative phase .
motivated by the observation of such soliton trains , a `` soliton laser '' has been proposed @xcite .
a review article on bec solitons addresses some of this work in more detail @xcite .
as opposed to solitary - wave trains , we investigate whether classic soliton - like behaviour , i.e. , colliding and reforming of distinct , localized wave packets up to shifts in phase and position , is possible in the 1d gpe with a harmonic potential . in a previous work @xcite
we found regimes where such behaviour is quite pronounced .
this behaviour was also seen in work done in similar nonlinear optical settings @xcite . in this paper
we further our investigation into soliton - like behaviour ; in particular we explore the bounds within which the solitary - waves can still be expected to behave as solitons . to this end
, we use a particle model introduced in our previous work @xcite , adapted from a model developed for use in nonlinear optics @xcite .
we show that soliton - like behaviour is possible in the 1d gpe with a harmonic potential , provided that the solitary - waves collide with large enough relative velocity such that the collisions occur during a short timescale compared with the period of the axial trapping potential . this type of behaviour has recently been experimentally observed @xcite , and provides an exciting prospect for future experiments to probe the dynamics in more detail . in the case of three solitons
, we find regimes of regular and chaotic dynamics .
in particular , chaotic solutions to the gpe are expected to coincide with more rapid condensate depletion than in otherwise similar regular solutions @xcite ; indeed this has been seen in theoretical studies of several systems @xcite .
this provides an additional motivation to identify regimes of regular and chaotic soliton dynamics in the gpe . in more realistic models for becs
, the integrability of the nonlinear wave equation is also broken by residual 3d effects .
these effects cause the soliton collisions to be inelastic ; specifically , there is particle exchange between the solitons accompanied by changes in their outgoing velocities @xcite .
a reduction from 3d to non - integrable 1d equations @xcite , more sophisticated than the 1d gpe , confirms this result @xcite .
this type of behaviour is common in other non - integrable schrdinger - type equations : @xcite .
bose - einstein condensates with attractive interatomic interactions are prone to collapse if the particle density becomes too high @xcite .
fully 3d gpe simulations show that in - phase collisions between solitons , during which the particle density becomes large , can cause collapse of the condensate , @xcite . in this paper , as well as identifying regimes in which soliton - like behaviour can still occur despite the mild breaking of integrability by the harmonic potential , we also briefly discuss regimes where solitons are expected to survive 3d integrability breaking .
the layout of the paper is as follows : in section ii we introduce the model equations and reiterate the soliton solution to the homogeneous 1d gpe ; in section iii we identify one , two and three solitary - wave solutions to the 1d gpe with a harmonic potential and introduce a particle model to test their soliton nature ; in section iv we present our conclusions .
in the case of an atomic bose gas , the dynamics of the system may be described , in the heisenberg picture , by the time - evolution of the bosonic field operator @xmath1 . at the temperatures typically encountered in atomic bec experiments ( nk ) , for sufficiently dilute bose gases , atomic interactions are dominated by low energy two - body collisions . in this case
, the atom - atom interactions are characterised by one parameter : the @xmath0-wave scattering length , @xmath2 .
moreover , we may generally replace the true interaction potential by an effective contact interaction , subject to an appropriate renormalization procedure @xcite .
hence , we let @xmath3 where @xmath4 and @xmath5 is the particle mass of the species .
depending on the species , @xmath2 may be positive or negative , corresponding to an effective repulsive or attractive interaction . by exploiting a feshbach resonance
, it may also be tuned using an external magnetic field @xcite .
the many - body hamiltonian for the system in second - quantized form is then given by : @xmath6 \hat{\psi}(\mathbf{r}),\ ] ] where @xmath7 is the hamiltonian for a single particle in the external trapping potential .
if bose - einstein condensation has occurred , we may define the condensate mode as the eigenfunction @xmath8 of the single - body density matrix @xmath9 with the largest eigenvalue @xcite .
we are then free to partition the field operator into condensate and non - condensate parts @xcite : @xmath10 where @xmath11 annihilates a particle in mode @xmath8 , and @xmath12 is the field operator for modes orthogonal to the condensate . in the case of
a trapped , almost fully bose - condensed dilute atomic gas , the dynamics of the condensate mode @xmath8 are largely governed by the following gpe @xcite : @xmath13\psi(\mathbf{r}),\label{gp}\ ] ] where @xmath14 is the total number of particles in the condensate and @xmath8 is normalised to one .
we now consider a cylindrically symmetric ( cigar - shaped ) harmonic trapping potential : @xmath15,\ ] ] where @xmath16 .
we also explicitly assume @xmath17 ( attractive inter - particle interactions ) , and determine that eq .
( [ gp ] ) reduces to the following 1d equation ( see appendix [ app:3dto1d ] for a derivation ) : @xmath18 and @xmath19 in units of @xmath20 , with @xmath21 and @xmath22 equal to the axial frequency @xmath23 in our units of inverse time ( @xmath24 ) . it can be shown that linear instabilities in the gpe directly imply , in both bogoliubov @xcite and equivalent number - conserving linearised approaches @xcite , that the population of the non - condensate component may rapidly become significant . for this reason
we expect regimes where the gpe dynamics are chaotic to coincide with rapid depletion of the condensate .
the gpe is a norm - conserving equation , so the depletion will not show up in the gpe dynamics ; however , we may use chaos in gpe dynamics as an indicator of depletion of the condensate mode in a realistic system .
this motivates the identification of chaotic trajectories in the gpe dynamics , as we discussed in the introduction ( sec .
[ sec : introduction ] ) . in the case of no axial potential
[ equivalent to setting @xmath25 in eq .
( [ s2 ] ) ] , it is possible to find exact solutions of eq .
( [ s2 ] ) @xcite .
a straightforward interpretation is in terms of a scattering problem @xcite ; in the limit @xmath26 , the solutions take the form of an arbitrary number of well separated ( incoming ) solitons : @xmath27 e^{i v_{j}(x - q_{j})}e^{i(2\eta_{j}^{2}+v_{j}^{2}/2 ) t}e^{i\alpha_{0j}}. \label{sol_1}\ ] ] here @xmath28 is the position of the peak of the @xmath29th soliton ; @xmath30 is the peak position at @xmath31 ; @xmath32 is the phase for a single soliton ( i.e. , in the absence of collisions with other solitons ) at @xmath33 , @xmath31 ; @xmath34 is the soliton velocity and @xmath35 gives the relative size of the soliton .
our normalisation condition implies @xmath36 , where @xmath37 is the number of solitons present .
the solitons come together and collide , during which time the form of the solution is complicated and solitons are not individually defined . however , as @xmath38 ,
the outgoing solitons re - emerge from the collisions unscathed , taking the same asymptotic form [ eq . ( [ sol_1 ] ) ] , up to shifts in position and phase : @xmath39 and @xmath40 , where the position shift @xmath41 and phase - shift @xmath42 of the @xmath29th soliton are given by @xcite : @xmath43 . \label{shifts}\ ] ] the positive sign applies if the soliton is on the left prior to the collision with the @xmath44th soliton ( @xmath45 ) , otherwise the negative sign applies . note that these shifts are dependent on the solitons initial speeds @xmath34 , and effective masses @xmath35 only , not on their relative phase .
hz , the radial trap frequency is @xmath46 hz , atomic mass and scattering length of @xmath47li .
a unit of @xmath48 is hence equal to @xmath49 m.,width=332 ] as shown in appendix b , for any solution to the 1d harmonic gpe , there exist other solutions with the same density profile , undergoing arbitrary amplitude harmonic oscillations at the trap frequency .
in particular , for any stationary solution , there exist corresponding solutions with the same density profile , which oscillate with the trap frequency but remain otherwise unchanged .
hence , a single bright soliton in a harmonic trap experiences an overall simple harmonic motion without any manifestation of internal dynamics in the soliton s density profile .
the density profile and phase behaviour of a single soliton can be found by first considering the form of a stationary soliton , and then inferring the behaviour of the oscillating version .
the stationary soliton will be a solution to the eigenvalue problem : @xmath50\right\}$ ] , where @xmath51 is a real valued function and the real - valued number @xmath52 is an initial phase .
we expect the single stationary soliton solution to be the metastable `` ground '' state of the system @xcite , which may be determined numerically , for example by propagating eq .
( [ s2 ] ) in imaginary time @xcite .
the numerically determined density @xmath53 for a parameter regime consistent with the @xmath47li experiments of strecker _ et al_. is shown in figure [ anz ] , and is compared to a bright soliton solution of the homogeneous 1d gpe , and to the ground state of the 1d linear schrdinger equation with harmonic potential . as expected , the solution of the 1d gpe with a harmonic potential is spatially slightly compressed compared to the bright soliton solution of the homogeneous gpe .
these two solutions , however , are quite similar ( and can be made more similar as @xmath22 is progressively reduced ) , and are quite distinct from the gaussian solution produced by the linear schrdinger equation .
we will exploit this similarity later in the paper . as shown by the treatment in appendix [ app : harmonicallyoscillating ]
, an oscillating soliton solution takes the form : @xmath54\exp\left(i\left[-\mu t+\langle p(t)\rangle x - s(t)\right]\right),\label{coherent_state}\ ] ] where @xmath55 is the position expectation value of the atomic ensemble , @xmath56 is the momentum expectation value , i.e. , @xmath57 and @xmath58 @xmath59 and @xmath60 are the initial position and momentum expectation values , respectively , and @xmath61 removing the nonlinearity reproduces the result for a coherent state .
when the nonlinearity is present , however , the stationary eigenvalue , @xmath62 , is dependent on the norm of the soliton , unlike the case of the linear schrdinger equation . in this section ,
we present a simple model of multiple trapped solitons , treating each of the solitons as a classical particle .
we explore the case of two harmonically trapped solitons , and present results comparing the trajectories in the particle model with simulations of the wave dynamics in the gpe .
these results allow us to determine the range of initial conditions for which the particle model is a good description of the system .
recall from sec .
[ sec : background ] that , in a homogeneous system , the trajectories of solitons emerging from collisions with each other are independent of the relative phase of the incoming solitons .
the only effect of the relative phase of the solitons is on the form of the wavefunction ( peak or trough ) during the collision .
this is illustrated in figure [ fig_phase1 ] .
the phase - independence of the solitons incoming and outgoing trajectories allows a model to be formulated that treats the solitons as classical particles , each with only the positional degree of freedom ( rather than position and phase degrees of freedom used , for example , in @xcite ) .
this model was introduced by scharf and bishop in the context of nonlinear optics @xcite , which we have adapted for the purpose of modeling a quasi-1d harmonically trapped bec @xcite . to construct the particle model ,
we first consider the homogeneous solution before introducing the effects of the harmonic trap . following the approach in @xcite
, one can derive an effective inter - soliton potential ( see appendix [ app : deduction ] ) : @xmath63 , \label{v_interaction}\ ] ] which treats the solitons as particles of position @xmath64 and effective mass @xmath35 , the parameters used to describe the bright soliton solutions of eq .
( [ sol_1 ] ) .
this potential reproduces the asymptotic position shifts [ eq .
( [ shifts ] ) ] in the homogeneous gpe for the outgoing particle trajectories , i.e. , the position shifts as the solitons become infinitely far apart .
it yields accurate results when @xmath65 , where @xmath66 and @xmath67 are the soliton velocities , and which gives a lower limit for the relative velocity for which the particle model is applicable .
figure [ fig_phase1 ] shows the particle trajectories predicted by our model interaction potential [ eq . [ v_interaction ] ] superimposed on the density profile dynamics predicted by solution of the homogeneous 1d gpe . when modelling bec dynamics , an upper limit to the solitons relative velocity is also imposed , because the contact - interaction potential between atoms , used to derive the gpe , assumes low energy inter - atomic collisions , and may not be applicable to condensates with high relative approach speeds @xcite .
fortunately , recent experiments @xcite show that solitons are generated with similar sizes , such that their velocities may easily fall within our model s range of validity . in sec .
[ sec : singlesoliton ] we showed that independent trapped solitons oscillate harmonically . a more general method for deriving the approximate motion of solitons in an external potential of arbitrary form was found by scharf and bishop @xcite , and is outlined in appendix [ app : motion ] for completeness . to combine the effects of the external potential and the soliton collisions
, we use the homogeneous solution [ eq .
( [ sol_1 ] ) ] as an ansatz , so that the solitons are still characterised by the parameters @xmath64 and @xmath35 .
figure [ anz ] shows this to be a reasonable approximation .
the following hamiltonian : @xmath68 reproduces the harmonic motion of the solitons , keeping the interpretation of @xmath35 as effective masses ( see appendix [ app : motion ] ) .
we assume that the soliton - soliton interactions are not affected by the introduction of the ( relatively loose ) harmonic trap and construct the full hamiltonian by adding in the contributions from the interaction potentials : @xmath69 , \label{ham4 } \end{split}\ ] ] where @xmath37 is the number of solitons .
this approach is expected to be valid for regimes when the timescale of the soliton - soliton collisions is much less than the period of the harmonic trap , such that the effects of the harmonic trap are negligible during the collisions .
the limits of this approach are further explored in secs .
[ sec : presentation ] and [ sec : discussion ] . in the case of two solitons ( @xmath70 ) , it is useful to define the following independent coordinates : the centre - of - mass position @xmath71 and the relative position @xmath72 .
the hamiltonian [ eq .
( [ ham4 ] ) ] then takes the form : @xmath73 where @xmath74 is the momentum canonically conjugate to @xmath75 , and @xmath76 the momentum conjugate to @xmath77 .
the hamiltonian is now clearly seperable into two parts : the centre - of - mass energy @xmath78 ( dependent on @xmath79 and @xmath75 only ) , and the interaction energy @xmath80 ( dependent on @xmath81 and @xmath77 only ) .
there are thus two independent constants of the motion , @xmath78 and @xmath80 , as many as there are degrees of freedom .
hence , the particle model for two solitons is integrable and the dynamics must be completely regular @xcite . in the case where the solitons have identical effective masses ( @xmath82 ) , the hamiltonian [ eq .
( [ ham_2s ] ) ] reduces to @xmath83 figure [ fig_poin_2 ] shows four poincar surfaces of section ( or poincar sections ) for the two particle system ; sections of @xmath84 versus @xmath85 are shown for different surfaces of different total energy @xcite .
these poincar sections demonstrate the regular behaviour of the integrable two particle system , as all trajectories lie on invariant tori in the phase space of the system .
there are two distinct regimes observable in these poincar sections . in the lower regions of the sections ,
the centre - of - mass energy , @xmath78 , is large and positive ; in this case the interaction energy , @xmath80 , has a large negative contribution from the interaction term , and the solitons interact strongly .
it is seen in section [ sec : presentation ] that in this regime , there is rapid energy exchange between the solitons , such that the soliton with lower amplitude oscillations is driven by the other soliton , which itself becomes damped . in the upper regions of the sections ,
@xmath78 is less positive , and hence @xmath80 is less negative , so the energy exchange between the solitons occurs over a longer time period . going into and re - emerging from a collision in the case of @xmath86 ( dark line ) and @xmath87 ( light line ) .
zero is indicated by the dotted line .
the difference @xmath88 is equal to the particle trajectory @xmath85 minus the position of the peak of the left - hand soliton , or equivalently the position of the peak of the right - hand soliton minus the particle trajectory @xmath89 .
( b ) difference between the curves in figure ( a ) .
note that the separation between the trajectories in the particle model and gpe dynamics is always larger in the @xmath90-phase case .
, title="fig:",width=226 ] going into and re - emerging from a collision in the case of @xmath86 ( dark line ) and @xmath87 ( light line ) .
zero is indicated by the dotted line .
the difference @xmath88 is equal to the particle trajectory @xmath85 minus the position of the peak of the left - hand soliton , or equivalently the position of the peak of the right - hand soliton minus the particle trajectory @xmath89 .
( b ) difference between the curves in figure ( a ) .
note that the separation between the trajectories in the particle model and gpe dynamics is always larger in the @xmath90-phase case .
, title="fig:",width=226 ] in the poincar sections of fig .
[ fig_poin_2 ] , we highlighted a number of trajectories in white .
these trajectories are plotted in position space as a function of time , overlaying density plots of corresponding 1d gpe solutions , in figs .
[ fig_two_b][fig_two_d ] .
we do this to test how accurately the particle model represents the gpe dynamics for a range of initial conditions . in particular , the wave - dynamics involve a phase variable not accounted for in the particle model . as shown in fig .
[ fig_phase1 ] , the phase difference between the solitons has an observable effect on the solitons form during collisons , although as the solitons tend asymptotically apart , the solitons density dynamics are insensitive to this . two solitons with equal norms in a harmonic trap
have the same collisional form for all subsequent collisions ; i.e. , two solitons initially colliding with a phase difference @xmath91 , will have this phase difference for all subsequent collisions ( see appendix [ app : collisionalform ] ) .
this property allows results of gpe simulations with repeated in - phase and @xmath90 out - of - phase collisions to be compared for any trajectory in the two - particle model .
regimes with phase differences between zero and @xmath90 are not considered here , but will generally be expected to display behaviour intermediate between that of the zero and @xmath90 cases .
the trajectories displayed in fig .
[ fig_two_b ] correspond to the fixed point ( marked by a white star ) in the poincar section of fig .
[ fig_poin_2](b ) .
we observe significantly better agreement between the particle model and the wave dynamics for the in - phase case [ fig .
[ fig_two_b](a ) ] than for the out - of - phase case [ fig .
[ fig_two_b](b ) ] , where an obvious discrepancy between the particle and wave dynamics gradually accumulates .
subsequent collisions consistently occur slightly earlier in the particle model [ dynamics induced by the hamiltonian ( [ ham_2s_identical ] ) ] than for the solitons propagated by the gpe [ eq . ( [ s2 ] ) ] , so that by @xmath92 quite a noticeable shift of the particle dynamics has taken place .
for the zero phase case , a similar systematic discrepancy is just observable at large @xmath19 , in fig .
[ fig_two_b](a ) ; however , this discrepancy is very small and not readily apparent for most of the evolution .
we note that @xmath92 corresponds to 7 seconds using the parameters of ref .
@xcite , i.e. , substantially longer than most typical experimental time scales .
figure [ fig_two_c ] shows trajectories equivalent to those of fig .
[ fig_two_b ] , but with an additional centre - of - mass motion .
this corresponds to the upper of the two trajectories marked in white in the poincar section shown in fig .
[ fig_poin_2](c ) , where there is gradual energy exchange between the solitons .
we see the same discrepancy between particle and wave dynamics as was observed in fig .
[ fig_two_b ] .
this is to be expected , because , as shown in appendix [ app : harmonicallyoscillating ] , incorporating a centre - of - mass oscillation into a solution of the harmonic gpe simply causes the wave - function s density profile to oscillate , without any further effect on its overall evolution .
the trajectories in fig .
[ fig_two_c2 ] correspond to the lower trajectory illustrated in the poincar section of fig .
[ fig_poin_2](c ) , where there is rapid energy exchange between the solitons . in this regime , good agreement is found in the case of in - phase collisions .
for the case of out - of - phase collisions , there is some degree of agreement , but there are obvious discrepancies .
notably , the solitons in the gpe appear to remain a fixed distance apart , whereas the particle trajectories necessarily show repeated collisions .
the dynamics shown in fig . [ fig_two_d ] correspond to the fixed point ( marked by a white star ) in the poincar section of fig .
[ fig_poin_2](d ) . in this regime ,
the solitons collide with a significantly faster relative speed than in figs .
[ fig_two_b][fig_two_c2 ] , and the collision time is accordingly shorter compared to the trap period . in this regime
it is clear that the agreement between the particle model and the wave dynamics is much better than in the previous cases ( figures [ fig_two_b ] to [ fig_two_c2 ] ) , with the relative phase having much less of an observable influence . to explain the observed discrepancies between corresponding particle and wave evolutions , particularly when the solitons collide out - of - phase ( see figs .
[ fig_two_b ] and [ fig_two_c ] ) , we must consider the assumptions made while composing the effective particle hamiltonian [ eq .
( [ ham4 ] ) ] .
when constructing this hamiltonian , we stated that the collision time should be small compared with the trap period . to characterise the timescale of a collision ,
it is instructive to consider how rapidly the trajectories in the particle model converge , to those deduced from motion of the soliton peaks in the gpe , after a collision .
we consider again the evolutions depicted in fig .
[ fig_phase1](a ) and [ fig_phase1](b ) .
figure [ fig_deltap ] shows the differences in position between the peak of the left - hand soliton evolved by the gpe , and the corresponding trajectory in the particle model during a collision .
as there is no external potential , the particle trajectories are asymptotically exact as @xmath93 .
it is clear that , in the case of an in - phase collision , the convergence of the particle and wave trajectories is more rapid than for a @xmath90 out - of - phase collision . in a harmonic trap ,
subsequent to a collision , two solitons can only move a finite distance apart ( i.e. , _ not _ asymptotically far ) before moving together and colliding once more .
the dynamics of solitons colliding with a @xmath90 phase difference are therefore not expected to agree as well with the effective particle model dynamics , compared with solitons colliding in - phase .
this effect builds up over time because the phase difference is preserved for repeated collisions .
figure [ fig_deltap ] shows that , in the @xmath90 out - of - phase case , compared to the particle trajectory , the trajectory of the soliton peak tends to be further away from the point of collision , with the two trajectories converging asymptotically as @xmath19 tends to a time infinitely before and after the time of collision .
due to the harmonic potential , the particle trajectories consequently start their return to the centre of the trap at points closer to the centre , compared to the @xmath90 out - of - phase solitons .
this explains why subsequent collisions take place earlier for the particle trajectories than predicted by gpe dynamics , as observed in fig .
[ fig_two_b ] . for large approach speeds ( fig .
[ fig_two_d ] ) , however , the collision time is sufficiently small for both in - phase and out - of - phase cases , such that any discrepancy between the predictions of the gpe and the effective particle model is too small to be observed .
figure [ fig_two_c2 ] represents a regime where the solitons do not separate well between collisions .
the particle model might not be expected to apply well to such regimes ; nevertheless , reasonably good agreement is observed in fig .
[ fig_two_c2](a ) , with somewhat less good agreement ( as expected ) for the @xmath90 out - of - phase case shown in fig .
[ fig_two_c2](b ) .
the density distribution of figure [ fig_two_c2](b ) is essentially a continuous collision of two solitons with a @xmath90 phase - difference , and has the appearance of two individual wavepackets that never cross .
these are expected to be better described by alternative treatments , described by gordon @xcite in the absence of any external potential , and by gerdjikov _
@xcite in the case of harmonic confinement , or may alternatively be treated by perturbative approaches @xcite . in 3d simulations , collisions of solitons with a @xmath90 phase difference have been predicted to have some degree of immunity against collapse , as opposed to in - phase collisions @xcite .
it is clear from the above analysis that systems of two solitons of equal norm will be able to maintain a @xmath90 phase difference for all repeated collisions for any choice of initial positions and momenta , provided the right initial phase difference is chosen , and thus remain stable .
systems of two solitons with different norms have not been studied here , but are an area for future study ; we recall that the particle model is expected to remain valid for regimes of different norms as long as @xmath94 , as explained in sec .
[ sec : particlemodel ] .
whereas for two solitons , the particle model dynamics are always regular ( see sec . [
sec : particlemodel ] ) , in the case of three solitons ( @xmath95 ) , the situation is quite different .
a useful coordinate system for the three soliton system is to be found in the normal coordinates of the system for small displacements of the particles from the origin : the centre - of - mass position @xmath96 @xmath97 ( corresponding to the `` stretch '' mode ) , and @xmath98 ( corresponding to the `` asymmetric stretch '' ) .
the stretch modes are similar to those used to describe vibrational dynamics in a tri - atomic molecule @xcite ; as the system is constrained to 1d , however , there is no analogue of the molecular bending mode . using these coordinates
, the three - particle hamiltonian [ eq . ( [ ham4 ] ) ] takes the form : @xmath99 \\&+\frac{\omega^{2}}{2}\left[z_{t}^{2}(\eta_{1}+\eta_{2}+\eta_{3})+z_{c}^{2}\frac{\eta_{1}\eta_{2}+\eta_{2}\eta_{3}+4\eta_{1}\eta_{3}}{\eta_{1}+\eta_{2}+\eta_{3}}+z_{r}^{2}\frac{\eta_{1}\eta_{2}\eta_{3}}{\eta_{1}\eta_{2}+\eta_{2}\eta_{3}+4\eta_{1}\eta_{3}}\right ] \\ &
-2\eta_{1}\eta_{2}(\eta_{1}+\eta_{2})\mathrm{sech}^{2}\left[\frac{2\eta_{1}\eta_{2}}{\eta_{1}+\eta_{2}}\left(\frac{\eta_{2}\eta_{3}+2\eta_{1}\eta_{3}}{\eta_{1}\eta_{2}+\eta_{2}\eta_{3}+4\eta_{1}\eta_{2 } } \right)z_{r}+z_{c}\right ] \\&-2\eta_{1}\eta_{3}(\eta_{1}+\eta_{3})\mathrm{sech}^{2}\left[\frac{2\eta_{1}\eta_{3}}{\eta_{1}+\eta_{3}}\left(\frac{\eta_{2}\eta_{3}-2\eta_{1}\eta_{2}}{\eta_{1}\eta_{2}+\eta_{2}\eta_{3}+4\eta_{1}\eta_{2 } } \right)z_{r}+2z_{c}\right ] \\&-2\eta_{2}\eta_{3}(\eta_{2}+\eta_{3})\mathrm{sech}^{2}\left[\frac{2\eta_{2}\eta_{3}}{\eta_{2}+\eta_{3}}\left(\frac{-(\eta_{1}\eta_{2}+2\eta_{1}\eta_{3})}{\eta_{1}\eta_{2}+\eta_{2}\eta_{3}+4\eta_{1}\eta_{2 } } \right)z_{r}+z_{c}\right ] , \label{ham_three } \end{split}\ ] ] where @xmath100 , @xmath101/(\eta_{1}+\eta_{2}+\eta_{3})$ ] , and @xmath102 are the momenta canonically conjugate to the coordinates @xmath103 , @xmath104 , and @xmath105 , respectively .
it is apparent in eq .
( [ ham_three ] ) that the hamiltonian , as in the two - particle case , is decoupled into a centre - of - mass component , and a component describing the stretch modes ( which are coupled to each other ) . in the case of
identical effective masses , the coordinates simplify substantially .
it turns out to be convenient to consider slightly different coordinates , however , as this produces a simpler final form for the hamiltonian describing the stretch mode dynamics .
we therefore define @xmath106 , @xmath107 and @xmath108 .
we rescale the time to @xmath109 , and then introduce the momenta @xmath110 and @xmath111 . using these dynamical variables ,
the resultant hamiltonian ( the reduced system hamiltonian ) , with the centre - of - mass degrees of freedom removed , becomes : @xmath112 this hamiltonian , describing the two remaining degrees of freedom , is not separable , and it is necessary to integrate the corresponding hamilton s equations of motion numerically to analyse the system s behaviour . as they represent a slice through the phase space of a system , poincar sections provide a good illustration of regions of regular and chaotic dynamics . in regions of regular behaviour , any trajectory will lie on a torus in phase - space , and will thus trace a closed curve in the poincar section ; in regions of chaotic behaviour , a trajectory will go through every point in that region of phase space , and thus fill an area on the poincar section ( a so - called ergodic sea ) @xcite .
we choose to show sections corresponding to the momentum @xmath113 and position @xmath114 of the `` asymmetric stretch '' mode when the `` stretch '' mode coordinate takes the value @xmath115 , and when its canonically conjugate momentum @xmath116 .
other sections can be expected to be equally illustrative of the qualitative behaviour .
figure [ fig_poin ] shows six poincar sections for three different reduced system energies @xmath117 .
the behaviour is regular at large positive values of @xmath117 [ fig .
[ fig_poin ] ( a ) ] , but as @xmath117 is reduced , chaotic behaviour emerges , characterised by ergodic regions in between regular tori . for small ( negative ) @xmath117 the system is mostly an ergodic sea , with islands of stability [ fig .
[ fig_poin ] ( c ) ] ; but as @xmath117 is made more negative , the chaotic regions begin to subside , and the behaviour becomes increasingly regular again .
consideration of the form of the reduced - system hamiltonian [ eq .
( [ ham_red ] ) ] shows that without the interaction the system is integrable , as it becomes a decoupled pair of harmonic oscillators .
when @xmath117 is large and positive , the interaction part of the hamiltonian ( which is always negative ) should give a relatively small contribution to the hamiltonian , compared to the integrable part of the hamiltonian ( which is always positive ) .
when @xmath117 is reduced , this is no longer the case , and chaotic dynamics are manifest .
however , in regimes where the coordinates and momenta are close to zero , i.e. , @xmath117 approaches its lower bound of @xmath118 , the interaction potential becomes approximately harmonic .
the hamiltonian @xmath117 takes the following seperable form : @xmath119 i.e. , it again describes a pair of decoupled harmonic oscillators .
we consequently expect the phase - space structure to be qualitatively similar in the opposing limits of @xmath117 very large and positive , and @xmath117 large and negative . from figs .
[ fig_poin](a ) and [ fig_poin](f ) , we do indeed observe this to be the case .
figure [ fig_traj ] shows a comparison of trajectories in the particle model with results from integrations of the 1d gpe [ eq .
( [ s2 ] ) ] for the three - soliton system , where the solitons all have equal effective masses . as with the two - soliton case ( sec .
[ sec : twosolitons ] ) , the trajectories in the particle model gradually acquire a shift with respect to the trajectories traced out by the gpe wavefunction peaks . in figs .
[ fig_traj](a ) and [ fig_traj](b ) the overall shift indicates that subsequent collisions tend to take place sooner in the particle model than is predicted by the gpe evolution ; interestingly , in fig .
[ fig_traj](c ) we observe the opposite , however . as before , these shifts are caused by an accumulation of small errors , due to the fact that within a harmonic confining potential the individual solitons do not move asymptotically far from each other subsequent to collisions . in fig .
[ fig_traj](c ) there is the added complication that two of the solitons appear to have formed a `` bound state . ''
the comparisons illustrate the good agreement between the particle model and the 1d gpe in the regimes in which the particle model is valid , i.e. , when solitons are well separated between collisions [ fig .
[ fig_traj](a ) and [ fig_traj](b ) ] , even when the motion is chaotic [ fig .
[ fig_traj](b ) ] .
when two of the solitons are not well separated [ fig .
[ fig_traj](c ) ] , the 1d gpe simulation shows that a `` bound state '' is formed , which looks like a single `` higher - order '' soliton with an excited breathing mode @xcite .
the particle model does not predict well the behaviour within the `` bound state '' , but does give a good prediction of the centre - of - mass motion of the `` bound state '' and its interactions with the other soliton ; it is likely that the behaviour of the density of the `` bound state '' is strongly coupled to the phase behaviour within the `` bound state . ''
as in the two soliton case ( sec . [
sec : twosolitons ] ) , errors gradually accumulate in the particle model which lead to an overall `` time shift '' in the overall collision dynamics . it should be noted , however , that apart from this shift qualitative agreement with the dynamics predicted by the gpe remains quite good right up until limits of our numerical calculations ( @xmath120 corresponding to 6 seconds for the experimental parameters in @xcite ) .
in this paper we have investigated the soliton - like nature of two and three bright solitary waves in harmonically trapped 1d bose - einstein condensates . to this end
we employed a model treating each soliton as a classical particle and compared the particle trajectories with simulations of the gross - pitaevskii equation .
the results from the particle model and the gpe display good agreement when the solitons collide with large relative velocities , such that the collision time is small with respect to the trap period .
when the solitons relative velocities are reduced , the trajectories in the particle model `` get ahead '' of those in the gpe simulations , i.e. , points on the particle trajectories can be identified with corresponding points in time during the gpe evolutions , but the gpe evolution is increasingly delayed by comparison ; this effect is more pronounced when the solitons collide out of phase , and is due to the non - zero time in which the particle trajectories asymptote to the wave trajectories after collisions .
we have shown that , when the external potential is harmonic , for systems of two solitons of equal size , the phase difference between the solitons is preserved for repeated collisions . in these systems ,
repeated out - of - phase collisions can cause the discrepancy between the dynamics in the particle model and the gpe to build up relatively rapidly .
we note that it is in this regime , i.e. , solitons colliding @xmath90 out of phase , that the solitons are predicted to be stable against collapse when 3d effects are considered .
finally , we have extended the treatment to systems of three harmonically trapped solitons . using the particle model we have shown
that , unlike the two soliton systems , systems of three solitons can display chaotic dynamics .
chaotic regimes rescind when the energy of the particle system ( with the centre - of - mass motion neglected ) is very positive or very negative . in these limits , the dynamics decouple to those of two separated simple harmonic oscillators .
the agreement between the dynamics in the particle model and the gpe simulations is good in both regular and chaotic regimes .
we thank j. brand , s. l. cornish , k .- p .
marzlin , t. s. monteiro , n. g. parker and n. r. walet for useful discussions , and acknowledge support from the uk epsrc .
the one - dimensional gpe can be obtained in approximation from the three - dimensional gpe by assuming a gaussian ansatz for the radial wavefunction and integrating out the radial dimensions .
the gaussian ansatz can be taken to be the approximate radial solution if the radial potential is sufficiently tight that the harmonic potential energy dominates over the interaction energy in the radial directions @xcite . beginning with the 3d gpe : @xmath121 \psi(\mathbf{r } )
\label{gpe1 } \end{split}\ ] ] where @xmath15,\ ] ] and @xmath8 must have unit norm .
we employ the ansatz @xmath122 , where @xmath123 is the harmonic oscillator ground state , with @xmath124 . averaging over the radial degrees of freedom @xmath125 , eq .
[ gpe1 ] becomes : @xmath126 \psi(x ) , \label{gpe2}\ ] ] where @xmath127 reassigning the zero of energy allows us to drop the constant term @xmath128 from the hamiltonian [ eq . ( [ gpe2 ] ) ] . introducing the dimensionless variables @xmath129 and @xmath130 eq .
( [ gpe2 ] ) then becomes : latexmath:[\[\frac{\partial}{\partial \tilde{t}}\tilde{\psi}(\tilde{x})=\frac{1}{2}\frac{\partial^{2}}{\partial \tilde{x}^{2}}\tilde{\psi}(\tilde{x})+\frac{\tilde{\omega}^{2}}{2}\tilde{x}^{2}\tilde{\psi}(\tilde{x})- we have assumed the @xmath0-wave scattering length @xmath2 in @xmath132 to be negative , and set @xmath133 and @xmath134 such that @xmath135 is normalised to one with respect to @xmath136 . in the case
where @xmath137 , eq .
( [ gpe1d_2 ] ) has the general soliton solution which takes the form , in the limit @xmath138 @xcite , @xmath139\bigr . \\ & \times \bigl.\exp\left[i v_{j}(\tilde{x}-v_{j}\tilde{t}-x_{0j})+(2\eta_{j}^{2}+v_{j}^{2}/2 ) \tilde{t}+i\alpha_{0j}\right]\bigr\ } \label{soliton } \end{split}\ ] ] the normalisation requirement assumed in eq .
( [ gpe1d_2 ] ) implies that @xmath140 . throughout the body of the paper , whilst the 1d system is being discussed , the tildes have been dropped for notational convenience .
here we show that an arbitrary solution to the 1d gpe with harmonic potential @xmath141\ ] ] can be converted to an oscillating solution with a modified phase : @xmath142\right\}.\ ] ] this general result is used twice in this paper ; firstly to infer the phase behaviour of a single soliton .
secondly it is used to show that any trajectory of two identical solitons is identical , in its collisional density , to the corresponding symmetrical case in which the solitons repeatedly collide in the centre of the potential .
we note in passing that this result holds for any nonlinear schrdinger equation with the nonlinearity given by some function of @xmath143 only .
if @xmath144 $ ] is a particular solution to the 1d gpe , where @xmath145 and @xmath146 are real - valued functions , the following coupled equations describe their behaviour : @xmath147 \label{wp_1}\ ] ] @xmath148 defining the functions @xmath149 and @xmath150 , where @xmath151 , and defining the coordinate @xmath152 the following relations follow trivially : @xmath153 @xmath154 @xmath155 @xmath156 also , the relationship between the partial derivatives of @xmath157 and @xmath146 is the same as that between @xmath158 and @xmath145 . from now on it is assumed that @xmath159 , @xmath160 and their derivatives are be functions of @xmath161 and @xmath19 .
( [ wp_1 ] ) and ( [ wp_2 ] ) become : @xmath162\ ] ] and @xmath163 by choosing @xmath164 such that @xmath165 and @xmath166 where @xmath167 we can relabel @xmath161 as @xmath48 , and find that @xmath168 and @xmath169 are solutions of eqs .
( [ wp_1 ] ) and ( [ wp_2 ] ) .
hence , @xmath169 has the profile of @xmath145 , but undergoes additional global harmonic oscillations at the trap frequency , and the result is proved .
the motion of a single soliton can be derived by adapting the method of scharf and bishop @xcite where a solution to the homogeneous gpe is used as an ansatz for the gpe with an external potential .
this method has the advantage that it can be used with other than harmonic potentials .
here we show that for a harmonic potential it confirms the expected result of simple harmonic motion . in the 1d gpe ,
the norm @xmath170 , and energy @xmath171 , are conserved quantities .
the single soliton solution of the homogeneous case [ eq . ( [ sol_1 ] ) with @xmath172 is used as an ansatz for the harmonic case .
evaluating the norm and energy functionals with this ansatz yields : @xmath173 @xmath174 \\=&2\eta\dot{q}^{2}-\frac{16}{5}\eta^{3}+\frac{\omega^{2}}{2}\left(4\eta q^{2}+\frac{\pi^{2}}{12\eta}\right ) .
\end{split}\ ] ] the conservation of the norm @xmath170 , leads to @xmath175constant , and the conservation of the energy @xmath171 , leads to an equation of motion for the peak of the soliton , @xmath77 : @xmath176
in order to deduce an effective inter - particle potential , it is necessary to perform a classical inverse - scattering calculation to find the potential which produces the same asymptotic position - shifts as given in ref .
@xcite for solitons emerging from a collision .
this proof follows the method given in ref .
@xcite .
the position - shifts given in ref .
@xcite are equivalent to the asymptotic time - shifts for two solitons of initial speeds @xmath66 and @xmath67 , and effective masses @xmath177 and @xmath178 , given by @xmath179 .
\label{time_wave}\ ] ] we wish to produce these shifts in a system of two classical particles described by the hamiltonian : @xmath180 where @xmath72 is the relative coordinate , @xmath181 is the reduced effective mass , @xmath182 is the relative momentum , and the centre of mass has been separated from the problem . for particles initially separated at infinity , and noting that @xmath183 , this hamiltonian takes the asymptotic form @xmath184 i.e. , we assume the potential must vanish asymptotically . by rearranging eq .
( [ h_scatter ] ) , we may write the infinitesimal @xmath185}{\mu } } dt,\ ] ] since energy is conserved over the whole trajectory . integrating the time difference between trajectories with and without the inter - particle potential , we determine the asymptotic timeshift to be @xmath186 .
\label{time_particle}\ ] ] now , expanding eqs .
( [ time_wave ] ) and ( [ time_particle ] ) in terms of powers of @xmath187 , and equating equal powers , we obtain @xmath188^{n}-\left[\frac { 4\eta_{1}\eta_{2}\left(\eta_{1}-\eta_{2 } \right)^{2}}{\eta_{1}+\eta_{2 } } \right ] ^{n}\right\}\\ & \times\frac{(-1)^{n}2^{n-1}\left[(n-1)!\right]^{2}}{\left(2n-1 \right ) ! } \end{split}\label{v_int1}\ ] ] for all positive integers @xmath189 .
we now evaluate the integral of a candidate potential : @xmath190^{n}\\ = & \frac{1}{2}\left(\frac{1}{\eta_{1}}+\frac{1}{\eta_{2 } } \right)\left[4\eta_{1}\eta_{2}\left(\eta_{1}+\eta_{2}\right ) \right]^{n } \\&\times \frac{(-1)^{n}2^{n-1}\left[(n-1)!\right]^{2}}{\left(2n-1 \right ) ! } \end{split}\label{v_int2}\ ] ] comparing the expressions ( [ v_int1 ] ) and ( [ v_int2 ] ) it follows that the potential @xmath191 gives the correct time shift in the limit @xmath65 .
let us consider a general parity operator @xmath192 , such that @xmath193 .
we want to know whether , if @xmath194 , it continues to be the case that @xmath195 for all @xmath19 .
we take the time derivative of @xmath196 . noting that @xmath197 $ ] we deduce from the 1d gross - pitaevskii equation [ eq . ( [ s2 ] ) ] that @xmath198\psi(x , t ) \right\ } \\ = & \left [ -\frac{1}{2}\frac{\partial^{2}}{\partial x^{2 } } + \frac{\omega^{2}x^{2}}{2 } -|\chi(x , t)|^{2 } \right]\chi(x , t ) .
\end{split}\ ] ] hence , we see that the time - evolutions of @xmath199 and @xmath200 are governed by the same differential equation . if we also choose an initial condition such that @xmath201 , it must therefore follow that @xmath202 for all @xmath19 .
in other words , parity is conserved in the sense that an initially symmetric wave function will have that symmetry preserved throughout its subsequent dynamical evolution . in this paper ,
the most important consequence of this result is that a system of two identical solitons with equal and opposite velocities will repeatedly collide with the exact same collisional form ( e.g. , in phase or @xmath90 out of phase ) at the exact centre of the trapping potential . using the results of appendix [ app : harmonicallyoscillating ]
, it follows that an equivalent result holds upon the addition of a centre of mass oscillation . | systems of solitary - waves in the 1d gross - pitaevskii equation , which models a trapped atomic bose - einstein condensate , are investigated theoretically . to analyse the soliton - like nature of these solitary - waves , a particle analogy for the solitary - waves
is formulated .
exact soliton solutions exist in the absence of an external trapping potential , which behave in a particle - like manner , and we find the particle analogy we employ to be a good model also when a harmonic trapping potential is present . in the case of two solitons ,
the particle model is integrable , and the dynamics are completely regular .
the extension to three particles supports chaotic regimes .
the agreement between the particle model and the wave dynamics remains good even in chaotic regimes . in the case of a system of two solitary waves of equal norm ,
the solitons are shown to retain their phase difference for repeated collisions .
this implies that soliton - like regimes may be found in 3d geometries where solitary waves can be made to repeatedly collide out of phase , stabilising the condensate against collapse . |
a decade ago , deep h@xmath0 observations indicated that some disk galaxies can support limited star formation at their extreme outer edge ( e.g. ferguson et al .
galex imaging then surprisingly revealed that m 83 ( thilker et al . 2005 ) and ngc 4625 ( gil de paz et al .
2005 ) have extended uv disks ( xuv - disks ) unapparent in the distribution of hii regions .
we have since demonstrated that outer disk sf activity is commonplace , with @xmath1 1/3 of nearby s0-sm galaxies having discernible xuv - disk structure ( thilker et al .
2007 ) . for detailed information , see the review by gil de paz ( this volume ) or thilker et al .
( 2007 ) .
the relative lack of hii regions compared to uv clumps in the low sfr outer disk has been largely explained as a stochastic effect , tied to the very limited hii region lifetime compared to the time - scale for uv production ( boissier et al .
2007 ) . however , alternative contributing factors ( top - light imf , low density ism ) have yet to be ruled out and motivate our hst analysis .
hst acs uv
visible imaging of eight xuv - disk fields was obtained for m83 .
single locations in each of ngc 5055 ( fig . 1 ) and ngc 2090 are also being studied .
we observed in four band - passes ( f150lp , f435w , f606w , and f814w ) using the wfc and sbc .
optical observations of ngc 2090 were obtained using wfpc2 ( after the failure of acs / wfc ) .
hst resolves the xuv - disk sources into loosely clustered complexes of individual stars .
these complexes , likely evolved ob associations , are low mass ( @xmath2 m@xmath3 ) , intermediate age structures .
only very few hst detections are consistent with being zero - age upper - ms stars having mass @xmath4 15 m@xmath3 ( fig .
h@xmath0 emission is detected from complexes in which they are found .
observed association sizes vary from 100 pc to @xmath1 500 pc with significant internal sub - clustering .
the largest groupings may be blended associations .
cmds ( fig .
1 ) suggest multiple generations within larger complexes ( up to age of @xmath1 200 myr ) .
boissier , s. , et al .
2007 , apjs , 173 , 524 ferguson , a. , et al .
1998 , apj , 506 , l19 gil de paz , a. , et al .
2005 , apj , 627 , l29 thilker , d. a. , et al . 2005 , apj , 619 , l79 thilker , d. a. , et al .
2007 , apjs , 173 , 538 | we describe hst imaging of recent star formation complexes located in the extended uv disk ( xuv - disk ) component of ngc 5236 ( m 83 ) , ngc 5055 ( m 63 ) , and ngc 2090 .
photometry in four fuv
visible bands permits us to constrain the type of resolved stars and effective age of clusters , in addition to extinction .
the preliminary results given herein focus on cmd analysis and clustering properties in this unique star - forming environment . |
we address one of the fundamental problems in uncertainty quantification ( uq ) : the mapping of the probability distribution of a random variable through a nonlinear function .
let us assume that we are concerned with a specific physical or engineering model which is computationally expensive .
the model is defined by the map @xmath0 .
it takes a parameter @xmath1 as input , and produces an output @xmath2 , @xmath3 . in this paper
we restrict ourselves to a proof - of - principle one - dimensional case .
let us assume that @xmath1 is a random variable distributed with probability density function ( pdf ) @xmath4 .
the uncertainty quantification problem is the estimation of the pdf @xmath5 of the output variable @xmath6 , given @xmath4 .
formally , the problem can be simply cast as a coordinate transformation and one easily obtains @xmath7 where @xmath8 is the jacobian of @xmath9 .
the sum over all @xmath10 such that @xmath11 takes in account the possibility that @xmath12 may not be injective .
if the function @xmath12 is known exactly and invertible , eq.([py ] ) can be used straightforwardly to construct the pdf @xmath13 , but this is of course not the case when the mapping @xmath12 is computed via numerical simulations .
several techniques have been studied in the last couple of decades to tackle this problem .
generally , the techniques can be divided in two categories : intrusive and non - intrusive @xcite .
intrusive methods modify the original , _ deterministic _ , set of equations to account for the stochastic nature of the input ( random ) variables , hence eventually dealing with stochastic differential equations , and employing specific numerical techniques to solve them .
classical examples of intrusive methods are represented by polynomial chaos expansion @xcite , and stochastic galerkin methods @xcite .
on the other hand , the philosophy behind non - intrusive methods is to make use of the deterministic version of the model ( and the computer code that solves it ) as a black - box , which returns one deterministic output for any given input .
an arbitrary large number of solutions , obtained by sampling the input parameter space , can then be collected and analyzed in order to reconstruct the pdf @xmath13 .
the paradigm of non - intrusive methods is perhaps best represented by monte carlo ( mc ) methods @xcite : one can construct an ensemble of input parameters @xmath14 ( @xmath15 typically large ) distributed according to the pdf @xmath16 , run the corresponding ensemble of simulations @xmath17 , and process the outputs @xmath18 .
mc methods are probably the most robust of all the non - intrusive methods .
their main shortcoming is the slow convergence of the method , with a typical convergence rate proportional to @xmath19 . for many applications quasi - monte carlo ( qmc ) methods @xcite
are now preferred to mc methods , for their faster convergence rate .
in qmc the pseudo - random generator of samples is replaced by more uniform distributions , obtained through so - called quasi - random generators @xcite .
it is often said that mc and qmc do not suffer the ` curse of dimensionality'@xcite , in the sense that the convergence rate ( but not the actual error ! ) is not affected by the dimension @xmath20 of the input parameter space .
therefore , they represent the standard choice for large dimensional problems . on the other hand , when the dimension @xmath20 is not very large , collocation methods @xcite are usually more efficient . collocation methods recast an uq problem as an interpolation problem . in collocation methods ,
the function @xmath21 is sampled in a small ( compared to the mc approach ) number of points ( ` collocation points ' ) , and an interpolant is constructed to obtain an approximation of @xmath12 over the whole input parameter space , from which the pdf @xmath13 can be estimated .
the question then arises on how to effectively choose the collocation points .
recalling that every evaluation of the function @xmath12 amounts to performing an expensive simulation , the challenge resides in obtaining an accurate approximation of @xmath22 with the least number of collocation points .
+ as the name suggests , collocation methods are usually derived from classical quadrature rules @xcite .
the type of pdf @xmath23 can guide the choice of the optimal quadrature rule to be used ( i.e. , gauss - hermite for a gaussian probability , gauss - legendre for a uniform probability , etc .
@xcite ) . furthermore ,
because quadratures are associated with polynomial interpolation , it becomes natural to define a global interpolant in terms of a lagrange polynomial @xcite .
also , choosing the collocation points as the abscissas of a given quadrature rule makes sense particularly if one is only interested in the evaluation of the statistical moments of the pdf ( i.e. , mean , variance , etc . )
@xcite .
on the other hand , there are several applications where one is interested in the approximation of the full pdf @xmath22 .
for instance , when @xmath12 is narrowly peaked around two or more distinct values , its mean does not have any statistical meaning . in such cases
one can wonder whether a standard collocation method based on quadrature rules still represents the optimal choice , in the sense of the computational cost to obtain a given accuracy . from this perspective
, a downside of collocation methods is that the collocation points are chosen a priori , without making use of the knowledge of @xmath21 acquired at previous interpolation levels .
for instance , the clenshaw - curtis ( cc ) method uses a set of points that contains nested subset , in order to re - use all the previous computations , when the number of collocation points is increased .
however , since the abscissas are unevenly spaced and concentrated towards the edge of the domain ( this is typical of all quadrature rules , in order to overcome the runge phenomenon @xcite ) , it is likely that the majority of the performed simulations will not contribute significantly in achieving a better approximation of @xmath22 .
stated differently , one would like to employ a method where each new sampling point is chosen in such a way to result in the fastest convergence rate for the approximated @xmath22 , in contrast to a set of points defined a priori .
as a matter of fact , because the function @xmath12 is unknown , a certain number of simulations will always be redundant , in the sense that they will contribute very little to the convergence of @xmath5 .
the rationale for this work is to devise a method to minimize such a redundancy in the choice of sampling points while achieving fastest possible convergence of @xmath5 .
clearly , this suggests to devise a strategy that chooses collocation points _ adaptively _ , making use of the knowledge of the interpolant of @xmath21 , which becomes more and more accurate as more points are added .
a well known adaptive sampling algorithm is based on the calculation of the so - called hierarchical surplus ( * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* ; * ? ? ?
* see e.g ) .
this is defined as the difference , between two levels of refinement , in the solution obtained by the interpolant .
although this algorithm is quite robust , and it is especially efficient in detecting discontinuities , it has the obvious drawback that it can be prematurely terminated , whenever the interpolant happens to exactly pass through the true solution on a point where the hierarchical surplus is calculated , no matter how inaccurate the interpolant is in close - by regions ( see figure [ fig : hierarchical_example ] for an example ) .
the goal of this paper is to describe an alternative strategy for the adaptive selection of sampling points .
the objective in devising such strategy is to have a simple and robust set of rules for choosing the next sampling point .
the paper is concerned with a proof - of - principle demonstration of our new strategy , and we will focus here on one dimensional cases and on the case of uniform @xmath4 only , postponing the generalization to multiple dimensions to future work .
it is important to appreciate that the stated goal of this work is different from the traditional approach followed in the overwhelming majority of works that have presented sampling methods for uq in the literature .
indeed , it is standard to focus on the convergence of the nonlinear unknown function @xmath21 , trying to minimize the interpolation error on @xmath21 , for a given number of sampling points . on the other hand
, we will show that the convergence rates of @xmath21 and of its cumulative distribution function can be quite different .
our new strategy is designed to achieve the fastest convergence on the latter quantity , which is ultimately the observable quantity of an experiment .
the paper is organized as follows . in section 2
we define the mathematical methods used for the construction of the interpolant and show our adaptive strategy to choose a new collocation points . in section 3
we present some numerical examples and comparisons with the clenshaw - curtis collocation method , and the adaptive method based on hierarchical surplus .
finally , we draw our conclusions in section 4 .
in section 3 , we compare our method with the cc method , which is the standard appropriate collocation method for a uniform @xmath23 . here , we recall the basic properties of cc , for completeness .
the clenshaw - curtis ( cc ) quadrature rule uses the extrema of a chebyshev polynomial ( the so - called ` extrema plus end - points ' collocation points in @xcite ) as abscissas .
they are particularly appealing to be used as collocation points in uq , because a certain subset of them are nested .
specifically , they are defined , in the interval @xmath24 $ ] as : @xmath25 one can notice that the the set of @xmath26 points is fully contained in the set of @xmath27 points ( with @xmath28 an arbitrary integer , referred to as the level of the set ) .
in practice this means that one can construct a nested sequence of collocation points with @xmath29 , re - using all the previous evaluations of @xmath12 .
collocation points based on quadratures are optimal to calculate moments : on the left - hand side is a label , such that @xmath30 is the mean , @xmath31 is the variance , and so on . on the right - hand side it is an exponent . ] @xmath32 where we used the identity relation , @xmath33 it is known that integration by quadrature is very accurate ( for smooth enough integrand ) , and the moments can be readily evaluated , without the need to construct an interpolant : @xmath34 where the weights @xmath35 can be computed with standard techniques ( see , e.g. @xcite ) .
the interpolant for the cc method is the lagrange polynomial .
the hierarchical surplus algorithm is widely used for interpolation on sparse grids .
it is generally defined as the difference between the value of an interpolant at the current and previous interpolation levels @xcite .
the simplest algorithm prescribes a certain tolerance and looks for all the point at the new level where the hierarchical surplus is larger than the tolerance .
the new sampling points ( at the next level ) will be the neighbors of the points where this condition is met . in one - dimension ,
the algorithm is extremely simple because the neighbors are only two points , that one can define in such a way that cells are always halved . in this work
, we compare our new method with a slightly improved version of the hierarchical surplus algorithm .
the reason is because we do not want our comparisons to be dependent on the choice of an arbitrary tolerance level , and we want to be able to add new points two at the time .
hence , we define a new interpolation level by adding only the two neighbors of the point with the largest hierarchical surplus .
all the previous hierarchical surpluses that have been calculated , but for which new points have not been added yet are kept .
the pseudo - code of the algorithm follows .
the interpolant is understood to be piece - wise linear interpolation , and the grid is @xmath36 $ ] . calculate the interpolant on the grid @xmath37 .
define @xmath38 and add them on the grid calculate the interpolant on the new grid calculate the hierarchical surplus on the last two entries of @xmath39 and store them in the vector @xmath40 find the largest hierarchical surplus in @xmath40 , remove it from @xmath40 and remove the corresponding @xmath10 from @xmath41 append the two neighbors to @xmath41 and add them to the grid we use a multiquadric biharmonic radial basis function ( rbf ) with respect to a set of points @xmath42 , with @xmath43 , defined as : @xmath44 where @xmath45 are free parameters ( referred to as shape parameters ) . the function @xmath21 is approximated by the interpolant @xmath46 defined as @xmath47 the weights @xmath48 are obtained by imposing that @xmath49 for each sampling point in the set , namely the interpolation error is null at the sampling points .
this results in solving a linear system for @xmath50 of the form @xmath51 , with @xmath52 a real symmetric @xmath53 matrix .
we note that , by construction , the linear system will become more and more ill - conditioned with increasing @xmath15 , for fixed values of @xmath54 .
this can be easily understood because when two points become closer and closer the corresponding two rows in the matrix @xmath52 become less and less linearly independent . to overcome this problem one needs to decrease the corresponding values of @xmath54 . in turns
, this means that the interpolant @xmath46 will tend to a piece - wise linear interpolant for increasingly large @xmath15 .
we focus , as the main diagnostic of our method , on the cumulative distribution function ( cdf ) @xmath55 , which is defined as @xmath56 where @xmath57 . as it is well known , the interpretation of the cumulative distribution function is that , for a given value @xmath58 , @xmath59 is the probability that @xmath60 .
of course , the cdf @xmath55 contains all the statistical information needed to calculate any moment of the distribution , and can return the probability density function @xmath13 , upon differentiation .
moreover , the cdf is always well defined between 0 and 1 .
the following two straightforward considerations will guide the design of our _ adaptive selection strategy_. a first crucial point , already evident from eq .
( [ py ] ) , is whether or not @xmath21 is bijective .
when @xmath21 is bijective this translates to the cdf @xmath55 being continuous , while a non - bijective function @xmath21 produces a cdf @xmath55 which is discontinuous .
it follows that intervals in @xmath10 where @xmath21 is constant ( or nearly constant ) will map into a single value @xmath61 ( or a very small interval in @xmath62 ) where the cdf will be discontinuous ( or ` nearly ' discontinuous ) .
secondly , an interval in @xmath10 with a large first derivative of @xmath21 will produce a nearly flat cdf @xmath55 .
this is again clear by noticing that the jacobian @xmath63 in eq .
( [ py ] ) ( @xmath64 in one dimension ) is in the denominator , and therefore the corresponding @xmath13 will be very small , resulting in a flat cdf @xmath55 .
+ loosely speaking one can then state that regions where @xmath21 is flat will produce large jumps in the cdf @xmath55 and , conversely , regions where the @xmath21 has large jumps will map in to a nearly flat cdf @xmath55 . from this simple considerations one can appreciate how important it is to have an interpolant that accurately capture both regions with _ very large _ and _ very small _ first derivative of @xmath21 . moreover , since the cdf @xmath55 is an integrated quantity , interpolation errors committed around a given @xmath62 will propagate in the cdf for all larger @xmath62 values .
for this reason , it is important to achieve a global convergence with interpolation errors that are of the same order of magnitude along the whole domain .
+ the adaptive section algorithm works as follows .
we work in the interval @xmath36 $ ] ( every other interval where the support of @xmath21 is defined can be rescaled to this interval ) . we denote with @xmath42 the sampling set which we assume is always sorted , such that @xmath65 .
we start with 3 points : @xmath66 , @xmath67 , @xmath68 . for the robustness and the simplicity of the implementation we choose to select a new sampling point always at equal distance between two existing points .
one can decide to limit the ratio between the largest and smallest distance between adjacent points : @xmath69 ( with @xmath70 ) , where @xmath71 is the distance between the points @xmath72 and @xmath73 .
this avoids to keep refining small intervals when large intervals might still be under - resolved , thus aiming for the above mentioned global convergence over the whole support . at each iteration
we create a list of possible new points , by halving every interval , excluding the points that would increase the value of @xmath74 above the maximum desired ( note that @xmath74 will always be a power of 2 ) .
we calculate the first derivative of @xmath46 at these points , and alternatively choose the point with largest / smallest derivative as the next sampling point .
notice that , by the definition of the interpolant , eq .
( [ interpolant ] ) , its first derivative can be calculated exactly as : @xmath75 without having to recompute the weights @xmath48 . at each iteration
the shape parameters @xmath45 are defined at each points , as @xmath76 , i.e. they are linearly rescaled with the smallest distance between the point @xmath73 and its neighbors .
the pseudo - code of the algorithm follows .
@xmath77 exclude points in @xmath78 such that @xmath79 calculate @xmath80 through ( [ first_derivative ] ) at @xmath81 alternatively choose @xmath78 with largest / smallest values of @xmath82 as new collocation point calculate new weights @xmath48
in this section we present and discuss four numerical examples where we apply our adaptive selection strategy . in this work we focus on a single input parameter and the case of constant probability @xmath83 in the interval
@xmath84 $ ] , and we compare our results against the clenshaw - curtis , and the hierarchical surplus methods . we denote with @xmath85 the interpolant obtained with a set of @xmath86 points ( hence the iterative procedure starts with @xmath87 ) . a possible way to construct the cdf @xmath55 from a given interpolant @xmath85 would be to generate a sample of points in the domain @xmath24 $ ] , randomly distributed according to the pdf @xmath16 , collecting the corresponding values calculated through eq .
( [ interpolant ] ) , and constructing their cdf . because here we work with a constant @xmath16 , it is more efficient to simply define a uniform grid in the domain @xmath88 $ ] where to compute @xmath85 . in the following
we will use , in the evaluation of the cdf @xmath55 , a grid in @xmath89 with @xmath90 points equally spaced in the interval @xmath91 $ ] , and a grid in @xmath92 with @xmath93 points equally spaced in the interval @xmath24 $ ] .
we define the following errors : @xmath94 where @xmath95 denotes the l@xmath96 norm .
it is important to realize that the accuracy of the numerically evaluated cdf @xmath55 will always depend on the binning of @xmath97 , i.e. the points at which the cdf is evaluated . as we will see in the following examples , the error @xmath98 saturates for large @xmath99 , which thus is an artifact of the finite bin size .
we emphasize that , differently from most of the previous literature , our strategy focuses on converging rapidly in @xmath100 , rather than in @xmath101 .
of course , a more accurate interpolant will always result in a more accurate cdf , however the relationship between a reduction in @xmath102 and a corresponding reduction in @xmath98 is not at all trivial .
this is because the relation between @xmath16 and @xmath13 is mediated by the jacobian of @xmath21 , and it also involves the bijectivity of @xmath12 .
+ finally , we study the convergence of the mean @xmath103 , see equation [ eq : mean ] , and the variance @xmath104 , which is defined as @xmath105 these will be calculated by quadrature for the cc methods , and with an integration via trapezoidal method for the adaptive methods .
+ we study two analytical test cases : * case 1 : @xmath106 ; * case 2 : @xmath107 ; and two test cases where an analytical solution is not available , and the reference @xmath21 will be calculated as an accurate numerical solution of a set of ordinary differential equations : * case 3 : lotka - volterra model ( predator - prey ) ; * case 4 : van der pol oscillator . while case 1 and 2 are more favorable to the cc method , because the functions are smooth and analytical , hence a polynomial interpolation is expected to produce accurate results , the latter two cases mimic applications of real interest , where the model does not produce analytical results , although @xmath21 might still be smooth ( at least piece - wise , in case 4 ) . in this case
@xmath21 is a bijective function , with one point ( @xmath108 ) where the first derivative vanishes .
figure [ fig : case_1_f ] shows the function @xmath21 ( top panel ) and the corresponding cdf @xmath55 ( bottom panel ) , which in this case can be derived analytically .
hence , we use the analytical expression of cdf @xmath55 to evaluate the error @xmath98 . the convergence of @xmath98 and @xmath102 is shown in figure [ fig : case_1_err ] ( top and bottom panels , respectively ) . here and in all the following figures blue squares denote the new adaptive selection method , red dots are for the cc methods , and black line is for the hierarchical surplus method .
we have run the cc method only for @xmath109 ( i.e. the points at which the collocation points are nested ) , but for a better graphical visualization the red dots are connected with straight lines .
one can notice that the error for the new adaptive method is consistently smaller than for the cc method . from the top panel , one can appreciate the saving in computer power that can be achieved with our new method .
although the difference with cc is not very large until @xmath110 , at @xmath111 there is an order of magnitude difference between the two .
it effectively means that in order to achieve the same error @xmath112 , the cc method would run at least twice the number of simulations .
the importance of focusing on the convergence of the cdf , rather than on the interpolant , is clear in comparing our method with the hierarchical surplus method .
for instance , for @xmath113 , the two methods have a comparable error @xmath101 , but our method has achieved almost an order of magnitude more accurate solution in @xmath55 . effectively , this means that our method has sampled the new points less redundantly . in this case
@xmath21 is an anti - symmetric function with zero mean .
hence , any method that chooses sampling points symmetrically distributed around zero would produce the correct first moment @xmath103 .
we show in figure [ fig : case_1_sigma ] the convergence of @xmath114 , as the absolute value of the different with the exact value @xmath115 , in logarithmic scale .
blue , red , and black lines represent the new adaptive method , the cc , and the hierarchical surplus methods , respectively ( where again for the cc , simulations are only performed where the red dots are shown ) .
the exact value is @xmath116 .
as we mentioned , the cc method is optimal to calculate moments , since it uses quadrature .
although in our method the error does not decrease monotonically , it is comparable with the result for cc . in this case
the function @xmath21 is periodic , and it presents , in the domain @xmath36 $ ] three local minima ( @xmath117 ) and three local maxima ( @xmath118 ) .
the function and the cdf @xmath55 are shown in figure [ fig : case_2_f ] ( top and bottom panel , respectively ) .
figure [ fig : case_2_err ] shows the error for this case ( from now on the same format of figure [ fig : case_1_err ] will be used ) .
the first consideration is that the hierarchical surplus method is the less accurate of the three .
second , @xmath102 is essentially the same for the cc and the new method , up to @xmath119 . for @xmath120
the cc methods achieve a much accurate solution as compared to the new adaptive method , whose error has a much slower convergence .
however , looking at the error in the cdf in top panel of figure [ fig : case_2_err ] , the two methods are essentially equivalent .
this example demonstrates that , in an uq framework , the primary goal in constructing a good interpolant should not be to minimize the error of the interpolant with respect to the true @xmath21 , but rather to achieve the fastest possible convergence on the cdf @xmath121 . although , the two effects are intuitively correlated , they are not into a linear relationship .
in other words , not all sample points in @xmath10 count equally in minimizing @xmath98 .
the convergence of @xmath122 ( exact value @xmath123 ) and @xmath114 ( exact value @xmath124 ) is shown in figures [ fig : case_2_mu ] and [ fig : case_2_sigma ] , respectively .
it is interesting to notice that our method presents errors that are always smaller than the cc method , although the errors degrade considerably in the regions between two cc points , where the two adaptive methods yield comparable results .
the lotka - volterra model @xcite is a well - studied model that exemplifies the interaction between two populations ( predators and preys ) .
this case is more realistic than cases 1 and 2 , as the solution of the model can not be written in analytical form .
as such , both the @xmath21 and the cdf @xmath55 used to compute the errors are calculated numerically .
we use the following simple model : @xmath125 where @xmath126 and @xmath127 denote the population size for each species ( say , horses and lions ) as function of time .
the ode is easily solved in matlab , with the ` ode45 ` routine , with an absolute tolerance set equal to @xmath128 .
we use , as initial conditions , @xmath129 , and we solve the equations for @xmath130 $ ] . clearly , the solution of the model depends on the input parameter @xmath10 .
we define our test function @xmath21 to be the result of the model for the @xmath131 population at time @xmath132 : @xmath133 the resulting function @xmath21 , and the computed cdf @xmath55 are shown in figure [ fig : case_3_f ] ( top and bottom panel , respectively ) .
we note that , although @xmath21 can not be expressed as an analytical function , it is still smooth , and hence it does not present particular difficulties in being approximated through a polynomial interpolant .
indeed the error @xmath102 undergoes a fast convergence both for the adaptive methods and for the cc method ( figure [ fig : case_3_err ] ) .
once again , the new adaptive method is much more powerful than the cc method in achieving a better convergence rate , and thus saving computational power , while the hierarchical surplus method is the worst of the three .
convergence of @xmath122 and @xmath114 are shown in figures [ fig : case_3_mu ] and [ fig : case_3_sigma ] , respectively .
similar to previous cases , the cc presents a monotonic convergence , while this is not the case for the adaptive methods . only for @xmath120 ,
the cc method yields much better results than the new method .
our last example is the celebrated van der pol oscillator@xcite , which has been extensively studied as a textbook case of a nonlinear dynamical system . in this respect
this test case is very relevant to uncertainty quantification , since real systems often exhibit a high degree of nonlinearity .
similar to case 3 , we define our test function @xmath21 as the output of a set of two odes , which we solve numerically with matlab .
the model for the van der pol oscillator is : @xmath134 the initial conditions are @xmath135 , @xmath136 .
the model is solved for time @xmath137 $ ] , and the function @xmath21 is defined as @xmath138 the so - called nonlinear damping parameter is rescaled such that for @xmath36 $ ] , it ranges between 50 and 250 .
the function @xmath21 and the corresponding cdf @xmath55 are shown in figure [ fig : case_4_f ] .
this function is clearly much more challenging than the previous ones .
it is divided in two branches , where it takes values @xmath139 and @xmath140 , and it presents discontinuities where it jumps from one branch to the other .
correspondingly , cdf @xmath55 presents a flat plateau for @xmath141 , which is the major challenge for both methods . in figure
[ fig : case_4_err ] we show the errors @xmath102 and @xmath98 .
the overall convergence rate of the cc and the new method is similar . for this case ,
the hierarchical surplus method yields a better convergence , but only for @xmath142 . as we commented before
, the mean @xmath122 has no statistical meaning in this case , because the output is divided into two separate regions . the convergence for @xmath114 is presented in figure [ fig : case_4_sigma ] .
we have presented a new adaptive algorithm for the selection of sampling points for non - intrusive stochastic collocation in uncertainty quantification ( uq ) .
the main idea is to use a radial basis function as interpolant , and to refine the grid on points where the interpolant presents large and small first derivative .
+ in this work we have focused on 1d and uniform probability @xmath16 , and we have shown four test cases , encompassing analytical and non - analytical smooth functions , which are prototype of a very wide class of functions . in all cases the new adaptive method improved the efficiency of both the ( non - adaptive ) clenshaw - curtis collocation method , and of the adaptive algorithm based on the calculation of the hierarchical surplus ( note that the method used in this paper is a slight improvement of the classical algorithm ) .
the strength of our method is the ability to select a new sampling point making full use of the interpolant resulting from all the previous evaluation of the function @xmath21 , thus seeking the most optimal convergence rate for the cdf @xmath55 .
we have shown that there is no one - to - one correspondence between a reduction in the interpolation error @xmath102 and a reduction in the cdf error @xmath98 .
for this reason , collocation methods that choose the distribution of sampling points a priori can perform poorly in attaining a fast convergence rate in @xmath98 , which is the main goal of uq .
moreover , in order to maintain the nestedness of the collocation points the cc method requires larger and larger number of simulations ( @xmath143 moving from level @xmath28 to level @xmath144 ) , which is in contrast with our new method where one can add one point at the time . + we envision many possible research directions to further investigate our method .
the most obvious is to study multi - dimensional problems .
we emphasize that the radial basis function is a mesh - free method and as such we anticipate that this will largely alleviate the curse of dimensionality that afflicts other collocation methods based on quadrature points ( however , see @xcite for methods related to the construction of sparse grids , which have the same aim ) .
moreover , it will be interesting to explore the versatility of rbf in what concerns the possibility of choosing an optimal shape parameter @xmath54 @xcite .
recent work @xcite investigated the role of the shape parameter @xmath54 in interpolating discontinuous functions , which might be very relevant in the context of uq , when the continuity of @xmath21 can not be assumed a priori .
finally , a very appealing research direction , would be to simultaneously exploit quasi - monte carlo and adaptive selection methods for extremely large dimension problems .
a. a. and c. r. are supported by fom project no .
12cser058 and 12pr304 , respectively .
we would like to remember dr.ir .
j.a.s . witteveen ( @xmath145 2015 ) for the useful discussions we had about uncertainty quantification .
43 natexlab#1#1[1]`#1 ` [ 2]#2 [ 1]#1 [ 1]http://dx.doi.org/#1 [ ] [ 1]pmid:#1 [ ] [ 2]#2 , , ( ) .
, , , ( ) . , , , , , , in : , , pp . . , , , ( ) .
, , , , ( ) . , , , , in : , gcms 11 , , , , pp . .
, , , ( ) . , , , .
, , , , . , , ( ) .
, , , , . , , , , .
, , , ( ) . , , ( ) .
, , , in : , , , pp . . , , , ( ) .
, , , ( ) .
, , , ( ) . , , , , ( ) .
, , , ( ) . , , ( ) .
, , ( ) . , , , ( ) .
, , , ( ) . , , , in : , , , pp . .
, , , ( ) .
, , ( ) . , , , ( ) .
, , in : , , , pp . . , ,
( ) . , , , ( ) . , , , , , , ( ) . , , , , ( )
, , , ( ) . , , , ( ) .
, , , ( ) . , , , ( ) .
( in black ) goes exactly through the red straight line at the points @xmath146 .
calculating the piece - wise linear interpolant between two ( @xmath147 ) , three ( @xmath148 ) , and five ( @xmath149 ) points would result in a null hierarchical surplus on these points.,width=10 ] ( top ) @xmath102 ( bottom ) as function of number of sampling points @xmath15 .
blue squares : new adaptive selection method .
red dots : clenshaw - curtis .
black curve : adaptive method based on hierarchical surplus.,width=10 ] ( top ) @xmath102 ( bottom ) as function of number of sampling points @xmath15 .
blue squares : new adaptive selection method .
red dots : clenshaw - curtis .
black curve : adaptive method based on hierarchical surplus.,width=10 ] ( top ) @xmath102 ( bottom ) as function of number of sampling points @xmath15 .
blue squares : new adaptive selection method .
red dots : clenshaw - curtis .
black curve : adaptive method based on hierarchical surplus.,width=10 ] ( top ) @xmath102 ( bottom ) as function of number of sampling points @xmath15 .
blue squares : new adaptive selection method .
red dots : clenshaw - curtis .
black curve : adaptive method based on hierarchical surplus.,width=10 ] | we present a simple and robust strategy for the selection of sampling points in uncertainty quantification .
the goal is to achieve the fastest possible convergence in the cumulative distribution function of a stochastic output of interest .
we assume that the output of interest is the outcome of a computationally expensive nonlinear mapping of an input random variable , whose probability density function is known .
we use a radial function basis to construct an accurate interpolant of the mapping .
this strategy enables adding new sampling points one at a time , _
adaptively_. this takes into full account the previous evaluations of the target nonlinear function .
we present comparisons with a stochastic collocation method based on the clenshaw - curtis quadrature rule , and with an adaptive method based on hierarchical surplus , showing that the new method often results in a large computational saving . |
in spite of the development of supercomputers and cutting - edge numerical methods , the simulation of experimental size many - body systems is still a hard task . following feynman s conjecture of a quantum simulator , it results useful to analyze manipulable systems able to reproduce , in a controllable way , the physical scenario that one wants to investigate .
the theoretical and experimental progresses of the last decades , boosted by nano - technological needs , made possible to combine the properties of cold gases and optical lattices to build - up artificial crystals able to mimic condensed - matter systems @xcite .
a crucial role in the success of these models is due to their experimental versatility .
interactions between cold atoms can indeed be tuned via feshbach @xcite , dipolar @xcite or confinement - induced @xcite resonances . on the other hand
optical lattices offer a full control of the potential landscape felt by cold atoms , allowing for the exploration of quantum phase transitions @xcite .
furthermore such kind of simulators enable to explore parameter ranges beyond those of the real material they imitate , unveiling new physical scenarios .
( depicted in blue ) and @xmath0 ( depicted in red ) are strongly confined on a plane . making use of a species - selective optical lattice @xcite @xmath0 atoms are arranged in a 2d lattice of point - like scatterers of arbitrary geometry ( a square one in this example ) while @xmath1 atoms form a matter wave which propagates through the artificial crystal.,scaledwidth=45.0% ] in solid state physics , among the plethora of crystals that can be investigated , two - dimensional ones are of special interest due to the intriguing properties that 2d materials have shown to posses . up to the early 2000s
the study of these systems was only of academic interest , 2d solids being considered unstable structures never observed experimentally .
things changed in 2004 , when graphene was finally isolated @xcite .
this discover paved the way to the study of this astonishing carbon allotrope constituted by a mono - layer of ions forming an honeycomb lattice in which charge carriers manifest peculiar transport properties @xcite . in particular conduction and
valence bands touch in isolated points of @xmath2-space : the dirac points . around them
the energy - momentum dispersion relation is conical and a dirac - like equation for massless fermions replaces schrdinger s one to describe the quantum motion of the carriers .
graphene qualifies thus as a quantum electrodynamics simulator on a benchtop scale .
furthermore relativistic effects , in general inversely proportional to the speed of light , would be enhanced in graphene : the role of @xmath3 is here played by the group velocity @xmath4 of the particles around the cone and @xmath5 @xcite .
the growing attention for graphene and others mono - layer materials translates in an increasing interest toward their quantum simulators , so that many artificial prototypes of 2d lattices have been proposed and realized in the past years ( for a recent review see @xcite ) . in the present work we introduce a general , highly controllable model for the realization of artificial bi - dimensional lattices , based on the use of two cold - atomic species . in our system
a 2d matter wave , made of @xmath1 atoms , interacts with point - like scatterers of a second atomic species , denoted by @xmath0 , independently trapped around the nodes of a 2d optical lattice . a schematic representation of our model is presented in fig .
[ figsystem ] .
such a scheme is already experimentally realizable using species - selective optical lattices : trapping potentials engineered to act on an atomic species ( @xmath0 in present case ) being at the same moment invisible to a second one ( for us @xmath1 ) @xcite .
this has been done , for instance , in @xcite for a mixture of @xmath6rb and @xmath7k atoms : tuning the optical - lattice frequency exactly in between two @xmath6rb resonances the attractive and repulsive contribution to the optical force cancel each other and only @xmath7k feels the added potential . in this study
we limit our investigations to one - body physics in the matter wave , i.e. we assume @xmath1 atoms to be non - interacting with themselves , situation attainable by using polarized fermions or bosons at zero scattering length .
one can instead employ the @xmath8 interaction to reach a mott insulating phase with exactly one atom per lattice site @xcite and subsequently freeze the atoms in this configuration by increasing the lattice depth .
other techniques are also available to probe @xcite and manipulate @xcite at single - site and single - atom level the scatterers arrangement .
this model has been recently proposed to study the effects of disorder in 1d @xcite , 2d and 3d systems @xcite , and it has the advantage to show a one - to - one correspondence with the bi - dimensional lattices that can mimic : the @xmath1 atoms of the matter wave play the role of the electronic cloud while the deeply trapped @xmath0 ones represent the crystalline structure . from now on we will refer to our system as atomic artificial crystal ( aac ) , since the periodic potential felt by the matter wave is generated by other atoms and not by a sub - standing optical potential .
the paper is organized as follows .
we start by introducing the theoretical model in sec .
[ sectheoreticalmodel ] , briefly discussing the problem of scattering in reduced _ and _ mixed dimensions ( sec .
[ secscattering ] ) and describing the general approach to the study of aac in sec .
[ secgeneral ] . in sec .
[ secbravais ] we specify the model to the case of ideal bravais arrangements of the scatterers , studying as illustrative examples the square ( sec . [ secsquare ] ) and triangular ( sec .
[ sectriangular ] ) lattices . for both of them the spectral properties
are analyzed in function of the @xmath9 interaction strength .
finite - size and disorder effects are also investigated .
[ secnonbravais ] is devoted to the generalization of previous results to non - bravais lattices , focusing on the exemplary atomic artificial graphene ( sec .
[ secgraphene ] ) @xcite and kagom lattice ( sec .
[ seckagome ] ) .
some properties of these systems , namely the emergence of dirac cones and non - dispersive flat bands , are pointed out and characterized .
we finally present our conclusions in sec .
[ secconclusions ] .
a remarkable feature of cold - atomic systems is the possibility to use experimentally controllable parameters as knobs to tune the interatomic scattering lengths , for example by means of feshbach or dipolar - induced resonances @xcite . when the interacting particles are subject to a trapping potential , this can , in turn , play a role leading to confinement - induced resonances @xcite .
if the trapping is sufficiently strong the dimensionality of the system can be reduced : it has been proved , for instance , that a 3d system subject to a strong axial confinement ( quasi-2d ) can be mapped into a strictly-2d one by introducing an effective 2d scattering length , the latter depending on the free 3d interaction and on the trapping parameters @xcite .
another related subject is the study of scattering processes involving differently - trapped atoms , in this cases we talk about scattering in mixed dimensions .
recently processes involving a free particle and a trapped one have been addressed ( nd-3d scattering @xcite ) and some theoretical predictions have been tested experimentally @xcite .
atom and a 0d - trapped @xmath0 one . the scattering process in the actual system ( top panel ) is ruled by the 3d @xmath9 scattering length @xmath10 of the contact interaction , as well as by the trapping frequencies and the mass ratio of the two atomic species .
the system is equivalent to a strictly bi - dimensional one ( bottom panel ) in which the @xmath0 atom constitutes a fixed point - like scatterer and the scattering process is ruled only by an effective 2d scattering length @xmath11 @xcite.,scaledwidth=45.0% ] in the case of the aac the building - block of any analysis is the 2-body low - energy scattering process between an @xmath1 atom of mass @xmath12 , harmonically trapped on a plane , and a @xmath0 one of mass @xmath13 , trapped around a node of an optical lattice , process schematically represented in fig .
[ figscattering ] .
it can be shown that , by providing a proper effective 2d scattering length @xmath11 , the system is mapped into a strictly 2d space in which the @xmath0 atom is now a fixed point - like scatterer .
the effective parameter @xmath11 , which takes also into account the quantum motion of the @xmath0 atom in the real system , depends on the atomic mass ratio , the trapping frequencies and the @xmath9 scattering length @xmath10 in free space ( c.f .
[ figscattering ] ) . a detailed study of this 0d-2d scattering process will be extensively presented elsewhere @xcite . for the purposes of present work we assume @xmath11 to be tunable in its full range of existence @xmath14 , protecting the validity of the point - like approximation for @xmath0 . in this regime
the interaction can be taken into account by considering the @xmath1 atom as free and imposing the bethe - peierls contact conditions
@xmath15{{{\bf r}}\to{{\bf r}}_{\!b } } \frac{m}{\pi\hbar^2 } d_{\!b } \ln\left(\frac{|{{\bf r}}-{{\bf r}}_{\!b}|}{{a_{2\rm d}^{\rm eff}}}\right ) + o\left(|{{\bf r}}-{{\bf r}}_{\!b}|\right)\ ] ] on its wave function at the position @xmath16 of the scatterer , where @xmath17 is an arbitrary complex coefficient .
we finally recall that , in the case of 2d scattering , the limit @xmath18 corresponds to a weakly attractive interaction , while for @xmath19 the effective potential allows a single , infinitely deep bound state and results weakly repulsive for positive - energy scattering states @xcite .
let us start by considering now a general case in which @xmath20 point - like @xmath0 scatterers are fixed at positions @xmath21 .
the first steps of our calculation follow @xcite : we consider the hamiltonian of the matter wave ( mw ) as that of a free @xmath1 atom , i.e. @xmath22 with @xmath23 the 2d laplace operator , and add the @xmath9 interaction by imposing the boundary conditions at the position of each @xmath0 scatterer .
this introduces a set of @xmath20 independent complex coefficient @xmath24 .
the same conditions apply to the mw green s function @xmath25 , solution of the schrdinger equation for a point - like source term of matter waves in @xmath26 : @xmath27 .
the latter wave equation can be rewritten to take directly into account the boundary conditions . by using the identity @xmath28 , the effect of contact conditions resumes in the inclusion of secondary point - like sources of amplitude @xmath24 at the position of each scatterer @xcite , leading to @xmath29 since the poles of @xmath30 ( and of its analytical continuation to complex energies in the lower half - plane ) correspond to eigenstates of the system ,
its knowledge is of fundamental importance to determine the properties of the aac .
to integrate eq .
we use its solution in absence of scatterers , that is the case of a free 2d mw . in this case
@xmath31 with @xmath32 where @xmath33 is the zero - index hankel function of the @xmath34 kind .
the wave vector modulus @xmath2 is linked to the mw energy by @xmath35 , with @xmath36 for @xmath37 and @xmath38 with @xmath39 for @xmath40 ( i.e. for bound states ) .
the formal solution of the wave equation is hence @xmath41 where the determination of the @xmath20 coefficients @xmath24 depends on the system geometry , encoded in the set @xmath21 .
this problem can be led back to the solution of a complex linear system in the @xmath20 unknowns @xmath24 .
each equation of such system comes from the limit @xmath42 of eq . , imposing the bethe - peierls condition on the lhs and applying @xmath43 { } 1+\frac{2i}{\pi}\ln\left(\frac{e^\gamma}{2}kr\right)+o(1)\ ] ] on the rhs , where @xmath44 is the euler - mascheroni constant . after some straightforward algebraic manipulation the system can be cast as @xmath45 with the introduction of the matrix @xmath46 of elements @xmath47 the formal solution of @xmath30 ( eq . )
has a pole if @xmath46 is not invertible , i.e. if @xmath48 .
the latter is thus our general condition for the existence of an eigenstate of the mw in the gas of scatterers .
the condition @xmath48 can be rewritten in a more practical form by noticing that the interaction - dependent terms appear only in the diagonal elements of @xmath46 . in particular
one can write @xmath49 with the introduction of the 2d interaction coefficient @xmath50 and for an arbitrary choice of the unitary length @xmath51 .
it follows that @xmath52 , for @xmath53 and @xmath54 the @xmath55 identity matrix . for @xmath40
the matrix @xmath56 is real and lookig for solutions of @xmath48 is equivalent to solve @xmath57 for each of the @xmath20 eigenvalues @xmath58 of @xmath56
. solutions of eq .
give real and negative energies of the mw bound states in the gas of scatterers .
for @xmath37 the situation is slightly different .
a continuum of states is allowed for the mw , nevertheless , for a large enough number of scatterers , precursors of the bulk bloch states of the infinite periodic system can be identified in the form of complex poles of the analytical continuation of @xmath30 to the lower - half plane of complex energies . in our approach , this corresponds to the fact that @xmath56 is now a complex matrix and the poles of the extended @xmath30 can be found by solving @xmath59 for complex energies of the form @xmath60 , where @xmath61 and @xmath62 represents , respectively , position and band - width ( i.e. inverse lifetime ) of the eigenstate .
the latter , in an extended and ordered system , would be a quasi - bloch state , i.e. a state showing the periodicity properties of a bloch one within the gas of scatterers but with a finite lifetime inside of it .
and @xmath63 . the shadowed area represents the unit cell .
right : reciprocal lattice in @xmath2-space corresponding to the square lattice .
consequently the reciprocal primitive vectors are @xmath64 and @xmath65 .
the @xmath34 brillouin zone ( fbz ) is shadowed .
the high - symmetry points @xmath66 , @xmath67 , and @xmath68 are highlighted and the @xmath69 path ( red , dashed ) constitutes an irreducible symmetry path.,scaledwidth=48.0% ] and @xmath70 .
the shadowed region represents the wigner - seitz unit cell .
right : reciprocal lattice for the triangular one , being @xmath71 and @xmath72 the reciprocal primitive vectors .
the fbz is shadowed and the high - symmetry points @xmath66 , @xmath73 , @xmath74 , and @xmath75 are highlighted .
the @xmath76 path ( red , dashed ) is an high - symmetry one.,scaledwidth=48.0% ] in this section we adapt the general formalism introduced in sec .
[ secgeneral ] to the case in which @xmath0 atoms are arranged in a bravais lattice : an infinite periodic structure where a unit cell , containing only _
one _ atom , is repeated to cover the entire 2d space ( c.f .
[ figschemesquare ] and [ figschemetriangular ] ) .
such a lattice is invariant under any translation @xmath77 with @xmath78 , where the set @xmath79 is defined in terms of the two primitive vectors @xmath80 and @xmath81 .
consequently we can define the reciprocal lattice as a periodic structure invariant under translations @xmath82 with @xmath83 , where @xmath84 and @xmath85 are the reciprocal primitive vectors , defined by the relation @xmath86 ( @xmath87)@xcite .
some examples of bravais and corresponding reciprocal lattices are presented in figs .
[ figschemesquare ] and [ figschemetriangular ] .
the condition for the existence of an eigenstate , i.e. @xmath48 , implies that the homogeneous system @xmath88 associated to the inhomogeneous one of eq .
admits a non - trivial solution .
notice that now the number of scatterers @xmath20 , and thus the number of equations and unknowns in the system , is infinite @xcite .
in such a periodic structure bloch s theorem holds , implying that @xmath89 where @xmath90 is a vector of the @xmath91 brillouin zone ( fbz ) in reciprocal space .
resorting to this property all the equations of the homogeneous system become identical , so that the unique condition to verify is @xmath92 where @xmath93 .
due to the slow convergence of the sum in eq .
, a rewriting of the equation in terms of reciprocal - lattice vectors is appropriate .
the delicate details of this transformation are reported in appendix [ appchangeofspacediag ] , the result being @xmath94 here @xmath95 is a coefficient depending only on the geometry of the bravais lattice , its origin and definition follow from the real - to - reciprocal lattice transformation , presented in app .
[ appchangeofspacediag ] .
we also reintroduced the 2d interaction coefficient @xmath50 , ( @xmath51 being an arbitrary unit of length ) and @xmath96 is the area of the real - space unit cell of the bravais lattice .
it is worth noticing that eq .
can be cast as @xmath97 with the introduction of the interaction - independent function @xmath98 it follows by its definition that @xmath99 diverges for @xmath100 where wave vector and energy of the mw are related by @xmath101 .
furthermore @xmath99 results monotonically decreasing in @xmath61 between two divergences , ensuring that for a given @xmath90 only _ one _ solution of eq
. exists between them .
another remarkable consequences is that if eq
. holds than condition is satisfied only for @xmath102 , that is in the limit of non interaction between mw and scatterers ( cf .
[ figdivergencesquare ] for a practical example ) .
this implies that for @xmath102 one would recover the dispersion relation of a free mw , whose energy is given by @xmath103 as a first example of aac we consider @xmath0 scaterers arranged in a square lattice of spacing @xmath51 , for which the primitive vectors are simply @xmath104 and @xmath63 . the reciprocal lattice and the fbz
are consequently defined , as illustrated in fig .
[ figschemesquare ] . ) , behavior of @xmath99 as a function of @xmath61 for different values of @xmath90 ( namely @xmath105 , @xmath106 , and @xmath107 from top to bottom ) . dashed , gray vertical lines represent the lowest values of @xmath108 for the selected @xmath90 . in the sketches on the right of each panel
a point indicates the corresponding @xmath90 inside the @xmath69 symmetry path . here
@xmath109.,scaledwidth=48.0% ] to study the band structure of the artificial atomic square lattice ( aasl ) one needs to solve eq .
, where in this case @xmath110 ( cf .
it results convenient from a computational point of view to evaluate once for all the function @xmath99 defined in eq . and
look afterwards for solutions of eq . for a given @xmath111 . in fig .
[ figdivergencesquare ] we plot @xmath99 for different values of @xmath90 inside the fbz . as expected @xmath112 diverges each time that @xmath113 ( i.e. when condition is satisfied ) , but when @xmath90 gets closer to the boundary of the fbz some values of @xmath108 corresponding to different @xmath114 eventually tend to each other .
a solution of eq . remains thus `` trapped '' in the small corridor formed by the two divergences and it tends to @xmath113 in the degeneracy limit , independently of @xmath111 . .
left : the four lowest energy bands are evaluated along the irreducible symmetry path @xmath69 within the @xmath34 brilouin zone ( see inset )
. dashed , gray lines correspond to the energy spectrum of a free matter wave .
right : density of states ( dos ) obtained by evaluating the energies of the same bands in @xmath115 points sampled inside the path . for each bin of the histogram ( of width @xmath116 , @xmath109 ) the band - normalized dos
is evaluated according to eq .
.,scaledwidth=48.0% ] we present in fig .
[ figspectersquare ] the spectrum and density of states ( dos ) of the aasl for @xmath117 .
the spectrum is evaluated along the @xmath69 symmetry path ( cf .
the free - mw energy @xmath108 is also shown and , as expected , only one solution of eq .
exists between two ( eventually coinciding ) free bands .
an omnidirectional gap is found in the band structure and , correspondingly , in the dos .
the numerical evaluation of the dos can be obtained by sampling each energy band in @xmath118 points within the fbz .
subsequently the formula @xmath119 gives the dos in the energy range @xmath120 , being @xmath121 the number of sampled energies falling in the energy interval .
the dos defined in eq . is normalized to the number of bands taken into account .
finally we remark that the lowest and isolated band lays entirely at negative energies for this value of @xmath111 , meaning that the mw bloch states are actually bound . .
the spectra are evaluated along the same path shown in fig .
[ figspectersquare ] . dashed , gray lines correspond to the energy spectrum of a free matter wave .
energies are normalized on @xmath109 .
, scaledwidth=48.0% ] a fundamental aspect of our model relies in the possibility to tune its physical properties by acting on @xmath111 .
this remarkable feature shows up in fig .
[ figspectercomparedsquare ] , where the spectra of the aasl are compared for different values of @xmath111 . the most evident modification concerns the lowest band , which results isolated for @xmath122 , so that a gap opens in the spectrum .
furthermore the band becomes more and more deep and flat , leading to small group velocities for the corresponding eigenstates .
for @xmath123 matter wave and scatterers are weakly interacting and , as expected , the band structure tends to the free - mw spectrum ( grey dashed curves in fig . [ figspectercomparedsquare ] ) .
$ ] for a system of @xmath124 scatterers arranged in a square lattice inside a disk of radius @xmath125 .
the energies are discretized with a step of @xmath126 ( @xmath109 ) .
for given @xmath40 all the @xmath20 solutions of eqs . are selected . for @xmath37
the sampled values are used as starting points to find @xmath20 complex poles of @xmath30 according to eq . .
only quasi - bloch bulk states are selected as explained in the text .
the color - map is applied to @xmath127 , where @xmath128 is the number of selected poles of @xmath30 within a rectangular bin of area @xmath129 ( @xmath130 and @xmath131 ) . @xmath132
indicate the positions of band boundaries as expected from the analysis of an infinite system ( figs .
[ figspectersquare ] and [ figspectercomparedsquare ] ) .
analogously the symbol @xmath133 marks the expected contact - point between the lowest and the first excited band , corresponding to the gap closure . ] up to now we considered ideal periodic systems , but for both theoretical and practical interests it is crucial to investigate the robustness of the above - reported results for realistic finite - size systems . in this case
one can not resort to bloch s theorem and the use of the general approach presented in sec .
[ secgeneral ] is necessary . in a typical experimental setup
, atomic clouds can be manipulated in optical lattices extending over @xmath134 sites per dimension .
this means that 2d artificial atomic lattices with @xmath135 trapped @xmath0-scatterers are experimentally realizable . in fig .
[ figfinitesizesquare ] we thus present the density of states on the @xmath136 $ ] plane for a finite system of @xmath137 scatterers arranged in a square lattice inside a circular region of radius @xmath125 .
the finite - size results are here compared with those of an ideal infinite system and the agreement is excellent . for @xmath138 and @xmath139 ( @xmath109 ) .
different panels correspond to different radii @xmath140 of the scattering region .
dot - dashed red vertical lines indicate in each panel @xmath141 , where @xmath142 is the average group velocity in the selected region of the @xmath136 $ ] plane.,scaledwidth=48.0% ] in the evaluation of the dos of fig .
[ figfinitesizesquare ] it results computationally convenient to fix @xmath61 and correspondingly look for values of @xmath111 satisfying conditions or .
for @xmath40 the @xmath55 real symmetric matrix @xmath143 and its eigenvalues can be evaluated , thus the @xmath111 solving eqs
. are immediately obtained . for @xmath37
one can use @xmath144 as starting point to find an approximate solution of the complex eqs .
we write @xmath145 and we assume the @xmath34 order expansion @xmath146 to be valid . by choosing @xmath147
$ ] each equation reduces to @xmath148+\delta z\,{m_i^o}'(\tilde{e})\!=\!0 $ ] , from which @xmath149 can be directly derived .
energies @xmath61 and band - width @xmath150 of the quasi - bloch states follow from @xmath151}{{m_i^o}'(\tilde{e})}.\ ] ] notice that the derivative @xmath152 can be evaluated resorting to the generalized hellmann - feynman theorem : for the complex symmetric matrix @xmath56 one has @xmath153\,\vec{u}_i$ ] , where @xmath154 is the right eigenvector of eigenvalue @xmath155 , normalized as @xmath156 ( instead of the usual @xmath157 ) . :
normalized density of bandwidths @xmath150 for all the positive - energy poles of the green function ( @xmath109 ) . the dot - dashed vertical blue line at @xmath158 marks the change in behavior of the density @xmath159 .
solutions with @xmath160 have been rejected because incompatible with the behavior of a quasi - bloch bulk state .
here we found @xmath161 ( i.e. @xmath162khz for a matter wave of @xmath6rb atoms in an aasl with @xmath163).,scaledwidth=42.0% ] the @xmath34 order approximation results sufficient if @xmath164 $ ] is small @xcite and this is the case for most of the solutions we find . in particular for quasi - bloch states
one expects a lifetime @xmath165 , where @xmath140 is the radius of the region containing the scatterers and @xmath4 the group velocity of the state . to verify this behavior we selected a small window on the @xmath136 $ ] plane , studying the distributions of @xmath166 when the system - size varies .
the results of this analysis are presented in fig .
[ figtaudistributionsquare ] .
we find that the distribution peak follows the expected behavior , confirming that the first - order approximation is sufficient to individuate quasi - bloch states .
nevertheless , when the poles of @xmath30 are evaluated by eq . , one can eventually find some results for which the @xmath34 order is insufficient .
these solutions would present non - physical negative values of @xmath150 and needs to be rejected ( in the case presented in fig .
[ figfinitesizesquare ] they constituted the 10.6% of positive - energy states ) .
furthermore , in a finite - size system , states others that quasi - bloch ones can appear ( such as edge - states ) for which the law @xmath167 is not valid . in order to tell them apart one can look at the density of @xmath150 , shown in fig .
[ figgammasquare ] for the same system considered in fig .
[ figfinitesizesquare ] .
a neat change in the behavior is found in @xmath168 , where @xmath169 is the highest estimated group velocity in the range of @xmath61 and @xmath111 considered .
states with @xmath160 can be in turn rejected , finally leaving only solutions behaving as quasi - bloch bulk states ( in the case of fig .
[ figfinitesizesquare ] this led to the exclusion of an additional 8.4% of solutions ) .
a natural question may arise concerning the dependence of @xmath169 on @xmath111 and @xmath61 .
it has been verified that by considering a cut - off depending on @xmath111 and @xmath61 results are qualitatively the same as those obtained using the aforementioned selection method .
for an aasl of radius @xmath170 , corresponding to 3209 available lattice sites .
different panels refers to different percentage of randomly occupied sites : 100% in ( a ) , 95% in ( b ) , 20% in ( c ) , 10% in ( d ) .
histograms are obtained with a bin size @xmath131 in ( a)-(b ) and @xmath171 in ( c)-(d ) . superimposed
dashed histograms in panels ( a)-(b ) represent the dos for an infinite periodic system .
the vertical lines in panels ( c)-(d ) indicate the energies of few - body bound states : @xmath172 dimer ( solid black ) , @xmath173 trimer with @xmath0 atoms separated by @xmath51 ( dashed blue ) and @xmath174 ( dot - dashed red ) . here
@xmath109.,scaledwidth=48.0% ] a remarkable advantage of our atomic artificial lattices with respect to other one - species models is the possibility to naturally introduce disorder in the system .
loading the @xmath0-atom optical lattice with non - unitary filling would result in the presence of randomly unoccupied sites constituting random defects in the artificial crystal @xcite .
our general theoretical approach , presented in sec .
[ secgeneral ] , consents to investigate the effects of this kind of disorder .
in particular we show in fig .
[ figdisordersquare ] the negative - energy dos of an aasl of radius @xmath170 at different filling factors , obtained fixing @xmath117 .
in panel ( a ) 100% of sites are occupied and finite - size effects can be investigated by comparing the dos with that of the periodic system ( already presented in fig . [ figspectersquare ] ) .
there are no significant discrepancies between the two quantities , confirming the robustness of the results with respect to the system size .
the dos in presence of 5% of vacant sites , panel ( b ) , appears in turns compatible with the results for an ideal periodic system , thus proving robustness also against a small amount of vacancies . for a large number of unoccupied sites , instead
, the periodicity of the lattice gets lost and the matter wave interacts locally with few - body clusters of scatterers . as can be seen in panels ( c ) and ( d ) of fig .
[ figdisordersquare ] , this gives rise to a dos which is more and more peaked around the energies of few - body @xmath175 bound states .
the energies of these bound states can be derived from eqs . , where the eigenvalues of @xmath56 can be analytically obtained for @xmath176 .
the explicit expressions can be found in eqs .
( 45 ) and ( 46 ) of @xcite , where the emergence of disorder - localized states for low filling have been investigated .
another relevant example of atomic artificial bravais lattice is the triangular one , which analysis is also preliminary to the study of intriguing non - bravais structures , such as graphene and the kagom lattice . in this case
the primitive vectors are @xmath177 and @xmath178 , from which the real and reciprocal lattice , illustrated in fig .
[ figschemetriangular ] , are defined . to investigate the spectral properties of the atomic artificial triangular lattice ( aatl ) we need , once again , to solve eq .
, now for @xmath179 .
we can thus proceed as done in sec .
[ secsquareinfinite ] for the square lattice . .
left : behavior of the four lowest energy bands along the @xmath76 path ( see inset to the right ) .
dashed gray lines indicates the dispersion relation for a free matter wave .
right : dos , as defined in eq . , obtained by evaluating the energy of the bands in @xmath180 points sampled inside the aforementioned symmetry path .
the bin - size of the histogram is @xmath131 ( @xmath109).,scaledwidth=48.0% ] .
the spectra - evaluation path is the same of fig .
[ figspectertriangular ] . dashed , gray lines correspond to the energy spectrum of a free matter wave .
energies are normalized on @xmath109.,scaledwidth=48.0% ] in fig .
[ figspectertriangular ] we present a typical spectrum of the system , evaluated in @xmath181 .
the band structure , studied along the @xmath76 high - symmetry path ( see inset ) , shows the presence of a gap , which the dos confirms to be omnidirectional .
also in this case we verify the existence of a single solution of eq . between two values of @xmath108 .
the versatility of this artificial lattice emerges in fig .
[ figspectercomparedtriangular ] , where we compare its spectra for different values of the interaction parameter @xmath111 .
one finds that there exists a gap of tunable width for @xmath182 and , again , the lowest and isolated band gets fast thin and deep in energy by decreasing @xmath111 . $ ] for a system of @xmath183 scatterers arranged in a triangular lattice inside a disk of radius @xmath184 . the same method of fig
. [ figfinitesizesquare ] has been used , but discretizing @xmath61 in steps of @xmath185 ( @xmath109 ) .
positive - energy quasi - bloch bulk states have been selected imposing @xmath186 ( i.e. @xmath187hz for a matter wave of @xmath6rb atoms in an aatl with @xmath163 ) .
the color - map is applied to the quantity @xmath127 , where @xmath128 is the number of selected poles of @xmath30 within a rectangular bin of area @xmath129 ( @xmath130 and @xmath171 ) . @xmath132
indicate the expected positions of the band boundaries as evaluated for the infinite system ( figs .
[ figspectertriangular ] and [ figspectercomparedtriangular ] ) .
analogously the @xmath133 marks the expected contact - point between the lowest and the first excited band.,scaledwidth=48.0% ] for sake of completeness we test again our results against finite - size effects . with the same method illustrated in sec .
[ secsquarefinite ] we evaluated the dos on the plane @xmath136 $ ] for an experimental - size system of @xmath188 scatterers arranged in a triangular lattice inside a circular region of radius @xmath184 .
the results , shown in fig . [ figfinitesizetriangular ] , are once again in good agreement with predictions based on the analysis of an ideal aatl .
we also verified that , in analogy to the square - lattice case , results are robust if a small amount of vacancies is randomly introduced in the triangular structure , while few body states becomes dominant for low fillings .
non - bravais lattice .
it can be seen as a triangular lattice ( fig .
[ figschemetriangular ] ) with 2 atoms per unit cell ( shaded hexagon ) whose relative position is @xmath189 .
equivalently it can be obtained as the superposition of two interpenetrating triangular lattices ( distinguished by colors ) displaced by @xmath189 with respect to each other .
the generating vectors @xmath190 are also indicated .
panel ( b ) : representation of the triangular lattice as example of basic bravais lattice , i.e. with _
one _ atom per unit cell .
the distance between nearest atoms is @xmath191 .
panel ( c ) : representation of the hexagonal lattice of graphene , a _ two_-atom non - bravais lattice based on the triangular one , obtained for @xmath192 . with these definitions ,
the side of the hexagons has length @xmath51 .
panel ( d ) : representation of the kagom lattice , a _ three_-atom non - bravais lattice based on the triangular one , obtained for @xmath193 and @xmath194 .
the nearest - neighbor distance results @xmath195.,scaledwidth=48.0% ] this section is devoted to generalize the formalism introduced in sec .
[ secbravais ] to the case in which @xmath0 scatterers are arranged in a non - bravais lattice : an infinite periodic structure in which a unit cell , containing now @xmath196 atoms , is repeated to cover the 2d plane .
such a structure can be equivalently seen as a set of @xmath196 identical bravais lattices , the @xmath197 and @xmath198 being displaced by @xmath199 with respect to each other
. the structure will be still invariant if translated by @xmath77 , for @xmath78 , where @xmath200 are the primitive vectors of a sub - lattice ( cf . panel ( a ) of fig .
[ figschemenb ] ) .
all the properties of a bravais lattice stay separately valid for each sub - lattice . if we denote by @xmath201 the central position of the @xmath202 unit cell and by @xmath203 the position of the @xmath197 atom with respect to this reference , the location of a scatterer in the non - bravais lattice is given by @xmath204 . the linear system becomes @xmath205 for each @xmath206 and @xmath207 . practically each index has been split with respect to previous notation ( i.e. @xmath208 and @xmath209 ) such that the first index runs on the lattice cells while the second indicates at which of the @xmath196 sub - lattices the scatterer belongs . accordingly the definition of @xmath46 given in eq . still holds .
bloch s theorem is now separately valid for each sub - lattice , so that @xmath210 while @xmath211 and @xmath212 stay independent for @xmath213 @xcite .
this assumption makes the present treatment not valid in the few accidental cases in which the non - bravais lattice degenerates in a bravais one . for an arbitrary choice of @xmath214 , all the equations of system take the form @xmath215 giving an homogeneous system of @xmath196 equations in the @xmath196 unknowns @xmath216 , whose matrix @xmath217 is defined by @xmath218
the matrix elements read explicitly @xmath219 with @xmath220 .
condition @xmath221 has to be satisfied so that system has a solution , which in turns means that this is the condition for the existence of an eigenstate of the @xmath1-atom mw in presence of a non - bravais lattice of @xmath0 scatterers .
notice that the diagonal terms are all the same and that they correspond to the lhs of eq . .
this naturally implies that for @xmath222 we get back to the case of a bravais lattice : @xmath223 is exactly eq . .
it is again convenient to re - write the sums in terms of reciprocal - lattice vectors . following the procedure described in app .
[ appchangeofspace ] we finally have @xmath224 besides , from the latter expressions the matrix @xmath217 results explicitly hermitian . considered the increasing interest towards the intriguing properties of graphene and its artificial realizations @xcite , it turns out natural to apply our model to investigate the behavior of the atomic artificial graphene ( aag ) ( cf .
@xcite for more details ) : a cold - atom quantum simulator of graphene where , differently from preexisting models presented in the field @xcite , the periodic potential felt by the matter wave is generated by other atoms .
the honeycomb structure is actually a non - bravais lattice resulting from the superposition of two interpenetrating triangular ones displaced by @xmath225 , as illustrated in fig .
[ figschemenb ] panel ( c ) .
( solid purple for @xmath133 and green dashed for @xmath226 ) as a function of @xmath61 for @xmath227 .
dashed , gray vertical lines indicates the lowest values of @xmath108 for the selected @xmath90 .
dot - dashed blue curves show the behavior of @xmath99 . in the inset
is shown the position of @xmath90 within the @xmath76 symmetry path ( fig .
[ figschemetriangular ] ) .
bottom : behavior of @xmath228 for the same @xmath90 . here
@xmath109.,scaledwidth=48.0% ] as usual we begin by considering an ideal infinite system . for a generic two - atom non - bravais lattice the condition for the existence of an eigenstate is @xmath221 , being the matrix @xmath217 a @xmath229 hermitian one .
once again the diagonal interaction term @xmath111 can be isolated by writing @xmath230 , where @xmath54 is the @xmath229 identity matrix and @xmath231 with @xmath99 defined as in eq . and @xmath232 it follows that condition @xmath221 is equivalent to @xmath233 where @xmath234 are the two real eigenvalues of @xmath235 .
it can be easily proved that when @xmath112 diverges , i.e. when @xmath236 , the absolute value of @xmath237 exactly cancels the divergence so that the left limit of @xmath238 is finite and equal to the right limit of @xmath239 : @xmath240 a numerical example of that is given in fig .
[ figdivergencegraphene ] , where the two eigenvalues @xmath241 and @xmath228 are plotted as functions of @xmath61 for a fixed @xmath90 . in virtue of the aforementioned properties , only _
one _ solution of eq .
exists between two solutions of the corresponding eq . .
this practically means that a band of a non - bravais lattice with @xmath242 is always included between two bands of the corresponding bravais one ( @xmath222 ) . .
left : behavior of the five lowest energy bands along the @xmath76 symmetry path ( see inset on the right ) .
dashed gray curves show the band structure for the corresponding triangular lattice .
right : dos ( defined in eq . ) obtained by sampling the bands energy in @xmath180 points within the symmetry path .
the histogram bin - size is @xmath131 ( @xmath109).,scaledwidth=48.0% ] an example of spectrum for the aag is presented in fig .
[ figspectergraphene ] for @xmath181 .
band structure and dos point out the presence of a double gap , which can be manipulated and eventually closed acting on @xmath111 .
a key feature is the existence of two nonequivalent dirac points , situated in @xmath243 within the fbz . here
the two lowest bands touch each other , creating a conic - shaped energy - momentum dispersion .
the matter - wave obeys so to a dirac - like equation for relativistic massless fermions in which the role of the speed of light is played by @xmath4 : the modulus of the group velocity of the wave along the cone .
remarkably the characteristics of the cones can be manipulated by acting on @xmath111 . at first for @xmath244
the cone lays at negative energies , so that relativistic physics is there played by bound states .
furthermore , similarly to the bravais artificial lattices , the lowest bands gets flat by decreasing @xmath111 leading to a widening of the cones ( cf .
[ figspectergrapheneflat ] ) .
this results in a sensitive decrease of @xmath4 , even below 1mm / s for typical experimental setups ( i.e. up to @xmath245 the corresponding quantity for electrons in real graphene ) .
evaluated along the @xmath76 symmetry path ( see inset of fig . [ figspectergraphene ] ) . for @xmath246
the third and isolated band results topologically flat . dashed gray curves show the band structure for the corresponding triangular lattices .
the unit of energy is @xmath109.,scaledwidth=48.0% ] another striking feature emerges in @xmath247 , values at which the third isolated band results topologically flat . as shown in fig .
[ figspectergrapheneflat ] , for @xmath248 the third - band maximum is in @xmath249 and the minimum in @xmath250 , while these roles are exchanged for @xmath251 .
correspondingly the band concavity changes at the transition point , leaving the band completely flat around @xmath252 ( @xmath109 ) .
it may be argued that also in the case of bravais artificial atomic lattices quasi - flat bands emerge for large and negative values of @xmath111 ( see lowest bands in figs .
[ figspectercomparedsquare ] and [ figspectercomparedtriangular ] ) , but some important differences exist between these bands and the one we find for the aag . at first graphene s band is topologically flat , which means that @xmath253 for every value of @xmath90 .
the quasi - flat bands keep instead their structure , even though they are compressed in a small range of energies ( this phenomenon appears clearly in fig .
[ figspectercomparedsquare ] ) . moreover graphene s flat band lies at positive energies , in the propagative region of the matter wave .
the interest towards such kind of flat bands arises from their non - dispersivity : for non - interacting @xmath1 atoms ( case considered in this work ) any mw state would result stationary and localized , being the group velocity strictly zero on the band .
the effects of an eventual @xmath254 interaction , even if extremely small , would be enhanced , leading to the emergence of strongly correlated phases @xcite .
non - isolated flat bands have been recently observed in honeycomb lattices for polaritons @xcite .
finally the use of an optical lattice allows to manipulate and distort the hexagonal structure , thanks to the precise experimental control achievable on the potential landscape . in one - component artificial graphenes such distortion of the structure is expected leading to the motion and eventual merging of the dirac cones @xcite , a phenomenon recently observed experimentally for a fermi gas in an honeycomb optical lattice @xcite .
similarly such a conductor - to - insulator transition is expected to occur in our atomic artificial graphene . in conclusion
the analysis of finite - size and disorder effects shows the robustness of the aforementioned results against the system size .
dirac cones and flat band persist for up to 2% of vacant sites . as an example of @xmath255 non - bravais lattice we consider the kagom one , motivated by an increasing interest toward this peculiar structure
in which several phenomena of geometric frustration have been predicted @xcite . like for graphene
, the kagom lattice is based on the triangular one , but now with _ three _ atoms per unit cell forming an equilateral triangle , as depicted in fig .
[ figschemenb ] panel ( d ) . evaluated along the @xmath76 symmetry path within the fbz ( see inset ) .
the two lowest bands touch in @xmath256 , forming a dirac cone .
for @xmath257 the fourth and isolated band results topologically flat .
for @xmath258 the three lowest band lay entirely at @xmath259 ( @xmath109 ) and are not shown in the plot .
dashed , gray lines correspond to the energy spectrum of a free matter wave.,scaledwidth=48.0% ] for the study of the periodic system , the condition @xmath221 translates now in looking for solutions of @xmath260 with @xmath261 and @xmath262 being the three eigenvalues of the matrix @xmath235 , defined in eq . .
typical spectra for different values of @xmath111 are presented in fig .
[ figspectercomparedkagome ] .
the two lowest bands present again two dirac cones for @xmath256 but , differently from the case of graphene , in the artificial kagom lattice the third band moves together with the lowest two .
the fourth band results flat and isolated for @xmath263 .
we notice again that for @xmath123 the band structure approaches that of the free mw , as expected being the @xmath9 interaction weak in this limit .
$ ] for a system of @xmath264 scatterers arranged in a kagom lattice inside a disk of radius @xmath265 .
results have been obtained with the same method described in fig .
[ figfinitesizesquare ] , discretizing the energies with a step of @xmath185 ( @xmath109 ) .
quasi - bloch bulk states have been selected choosing @xmath266 ( i.e. @xmath267hz for a matter wave of @xmath6rb atoms for @xmath163 ) .
the color - map is applied to the quantity @xmath127 , where @xmath128 is the number of selected poles of @xmath30 within a rectangular bin of area @xmath129 ( @xmath130 and @xmath171 ) .
@xmath268 , @xmath132 , and arrows indicate , respectively , the positions of dirac cones , gap boundaries , and isolated flat band as expected from the analysis of an infinite system ( figs .
[ figspectercomparedkagome]).,scaledwidth=48.0% ] the persistence of the spectral features in systems of experimental size have been also investigated . the density of solutions of eqs . - for a system with @xmath269 @xmath0 scatterers is shown in fig .
[ figfinitesizekagome ] .
the agreement with the predictions for an infinite system is extremely good , demonstrating once again the robustness of our results against finite - size effects .
in this work we presented a model for the realization of 2d arbitrary atomic artificial crystals based on the use of two independently trapped atomic species . this system revealed itself promising as a quantum simulator of bi - dimensional condensed - matter systems .
we discussed how the interplay between scattering length and trappings allows to widely tune the inter - species interaction : a key parameter trough which the system features can be manipulated .
a general theory for finite and infinite periodic systems has been presented , specifying to some illustrative examples for both bravais and non - bravais lattices .
we proved the emergence of single and multiple gaps , together with the eventual presence of isolated non - dispersive flat bands .
furthermore we pointed out the existence of dirac points in atomic artificial graphene and kagom lattices .
the robustness of our results for experimentally realizable systems has been tested against both finite size and disorder .
the adaptability of our model makes it suitable for a number of future developments .
new perspectives can be opened considering also the @xmath270-wave @xmath9 interaction @xcite or the @xmath254 one .
the latter , together with the occurrence of flat bands , would make the atomic artificial lattices convenient simulators of 2d strongly correlated systems in which non - trivial effects due to geometrical frustration may arise .
this appendix is devoted to the transformation of sums over real - lattice vectors , appearing in eqs.-- , in sums over reciprocal - lattice ones .
we manipulate here the lhs of eq . or , equivalently , a diagonal element of the matrix @xmath217 defined in eq . .
in particular we focus on the sum , rewriting @xmath271 , introduced in eq .
, as @xmath272 in terms of its fourier transform @xmath273 where @xmath274 denotes cauchy s principal value .
once sum and integral are exchanged we obtain @xmath275 where , by adding and subtracting the @xmath276 term , we have now a sum running over the entire @xmath79 .
for such a summation poisson s identity holds , stating that @xmath277 being @xmath96 the unit cell area in real space . in the case of interest
@xmath278 is the fourier transform of @xmath279 $ ] . by integrating out terms involving @xmath280-functions we obtain @xmath281
the principal - valued integral remains to be evaluated .
for an arbitrary choice of @xmath282 such that @xmath283 , we have @xmath284 and the bravais - element @xmath285 becomes @xmath286 we need now to introduce two auxiliary quantities : @xmath287 and @xmath288 where an arbitrary ultra - violet cut - off uv is added in the domains of sum and integration .
notice that the second line of eq .
is exactly @xmath289 , that is @xmath290 in the limit of a cut - off boundary pushed to infinity . from definitions and
it follows that @xmath291 { } - \frac{1}{2}\ln\!\left(\frac{k^2}{\rho^2}-1\right ) \nonumber\\ & + \frac{2\pi}{{\mathcal{a } } } \sum_{{{\bf k}}\in rl^ * } \left(\frac{1}{k^2-|{{\bf k}}-{{\bf q}}|^2}+\frac{1}{k^2}\right).\end{aligned}\ ] ] we thus add and subtract @xmath292 to eq . and
after some algebraic manipulation we obtain @xmath293 where we introduced the parameter @xmath50 and the arbitrary unit of length @xmath51 . from eq
. finally follows the definition of @xmath294 which numerically converges to a @xmath282-independent quantity determined only by the geometrical properties of the bravais lattice .
the transformation of the off - diagonal element of @xmath217 introduced in eq . follows straightforwardly from the previous one .
we write again @xmath271 in terms of its fourier transform @xmath295 , obtaining @xmath296 where , in the last step , we made use of poisson s identity as introduced in eq . . finally eq
. directly follows from eq . by writing explicitly @xmath297 .
m. marinescu and l. you , phys .
lett . * 81 * , 4596 ( 1998 ) ; z. y. shi , r. qi , and h. zhai , phys .
a * 85 * , 020702(r ) ( 2012 ) ; n. bartolo , d. j. papoular , l. barbiero , c. menotti , and a. recati , phys .
a * 88 * , 023603 ( 2013 ) .
g. montambaux , f. pichon , j .-
fuchs , and m. o. goerbig , eur .
j. b * 72 * , 509 ( 2009 ) ; l .- k .
n . fuchs , and g. montambaux , phys .
* 108 * , 175303 ( 2012 ) ; j. ibaez - azpiroz , a. eiguren , a. bergara , g. pettini , and m. modugno , phys .
a * 88 * , 033631 ( 2013 ) .
l. santos , m. a. baranov , j. i. cirac , h .- u .
everts , h. fehrmann , and m. lewenstein , phys .
lett . * 93 * , 030601 ( 2004 ) ; g .- b .
jo , j. guzman , c. k. thomas , p. hosur , a. vishwanath , and d. m. stamper - kurn , phys .
lett . * 108 * , 045305 ( 2012 ) . | we present a general scheme to realize a cold - atom quantum simulator of bi - dimensional atomic crystals .
our model is based on the use of two independently trapped atomic species : a first one , subject to a strong in - plane confinement , constitutes a 2d matter wave which interacts only with atoms of a second species , deeply trapped around the nodes of a 2d optical lattice . by introducing a general analytic approach
we show that the system green function can be exactly determined , allowing for the investigation of the matter - wave transport properties .
we propose some illustrative applications to both bravais ( square , triangular ) and non - bravais ( graphene , kagom ) lattices , studying both ideal periodic systems and experimental - size and disordered ones .
some remarkable spectral properties of these atomic artificial lattices are pointed out , such as the emergence of single and multiple gaps , flat bands and dirac cones .
all these features can be manipulated via the inter - species interaction , which results widely tunable due to the interplay between scattering length and confinements . |
zitterbewegung ( the trembling motion ) was theoretically devised by schroedinger @xcite after dirac had proposed his equation describing free relativistic electrons in vacuum .
schroedinger showed that , due to a non - commutativity of the quantum velocity @xmath2 with the dirac hamiltonian @xmath3 , relativistic electrons experience the zitterbewegung ( zb ) even in absence of external fields .
the frequency of zb is about @xmath4 and its amplitude is about the compton wavelength @xmath5 .
it was later understood that the phenomenon of zb is due to an interference of electron states with positive electron energies ( @xmath6 ) and those with negative energies ( @xmath7 ) , see @xcite . in other words , the zb results from the structure of the dirac hamiltonian , which contains both positive and negative electron energies , and it is a purely quantum effect as it goes beyond newton s first law .
an important step in the understanding of zb was made by foldy and wouthuysen @xcite , ( see also @xcite ) , who showed that in absence of external fields there exists a unitary transformation that transforms the dirac hamiltonian into a hamiltonian in which positive and negative electron energies are decoupled .
while solutions of the dirac equation are four - component functions , the transformed states for the positive energies have only two upper non - vanishing components and those for the negative energies have only two lower non - vanishing components . now
the above mentioned interference between the positive and negative energy states can not occur and there is no zb . instead ,
in the new representation the electron is not a point - like particle , but it acquires a quantum radius of the size @xmath8 .
the interpretation of the two pictures is until present not quite clear , see @xcite . to our knowledge , the zb for free electrons has never been directly observed .
however , in the presence of the coulomb potential the zb is manifested in appearance of the so called darwin term @xcite .
it was pointed out some time ago that the zitterbewegung also may occur in non - relativistic two - band systems in solids @xcite .
it was shown that , similarly to the relativistic case in vacuum discussed above , the consequence of the zb is that it is impossible to localize the electron better than to a certain finite volume .
recently , an analogy between the dirac description of electrons in vacuum and the coupled - band @xmath9 formalism for electrons in narrow - gap semiconductors ( ngs ) and carbon nanotubes ( cnt ) was used to demonstrate that the zb should occur in these systems @xcite .
it was shown that , in agreement with the semi - relativistic analogy @xcite , the zb frequency is always @xmath10 , where @xmath1 is the energy gap between the conduction and valence bands .
the amplitude of zitterbewegung in ngs and cnt was estimated to be @xmath11 , where @xmath12 is the effective electron mass and @xmath13cm / s is the maximum electron velocity in the system .
the zb length in ngs and cnt turns out be @xmath14 , i.e. @xmath15 times larger than in vacuum . a much lower zb frequency and its much higher amplitude , as compared to vacuum ,
should make the zb much more readily observable in semiconductors .
the zitterbewegung was also recently proposed in two - dimensional systems exhibiting spin splitting due to structure and bulk inversion asymmetry @xcite , and in 2d graphite @xcite .
a phenomenon similar to the zb was proposed for electrons in degenerate valence bands in the presence of an external electric field @xcite .
very recently , a unified description of the zitterbewegung of electrons in different solid state systems was attempted @xcite . in view of this
recently published work we want to investigate the question of whether the phenomenon of zitterbewegung in solids is a rule rather than an exception or _
vice versa_. to this end we consider two limiting models for electrons in solids : nearly - free electrons , for which the periodic potential of the lattice may be treated as a week perturbation , and tightly - bound electrons , for which the periodic potential may not be treated as a perturbation . since we are interested in general properties of the zitterbewegung , we do not insist on details of the band models in question but rather concentrate on essential features that result in this phenomenon .
although we deal with non - relativistic electrons in solids , we use methods of relativistic quantum mechanics to investigate an alternative picture in which the trembling motion is replaced by a kind of electron smearing in real space .
the reason , that a somewhat mysterious phenomenon of zitterbewegung of electrons in vacuum has never been observed , seems to be related to its very high frequency and very small amplitude .
the corresponding phenomenon in solids would have much lower frequency and much larger amplitude .
the underlying hope motivating our work is , that a more thorough theoretical understanding of the trembling motion will lead to an experimental detection of the phenomenon .
this would not only deepen our knowledge of electrons in solids but also represent a great success of the relativistic quantum theory .
our paper is organized in the following way . in section
ii we give the basis of nearly - free electron formalism , section iii treats the resulting zitterbewegung using schroedinger s method of the equation of motion . in section
iv a more realistic description of the zb is presented in which electrons are treated as wave packets . in section
v we use the foldy - wouthuysen transformation known from the relativistic quantum mechanics to obtain an alternative electron picture .
section vi treats the zitterbewegung in case of tightly bound electrons . in section vii
we discuss the obtained results and confront them with the previous work .
the paper is concluded by a summary .
the beginning of this section is standard , but it is needed for further developments .
we consider an electron in the presence of an external periodic potential @xmath16 , where @xmath17 is a translation vector of the lattice .
the periodic potential @xmath18 may be expressed by the fourier series @xmath19 , where @xmath20 are reciprocal lattice vectors and @xmath21 are fourier components of the periodic potential . for a real potential
there is @xmath22 .
the wave function of an electron has the bloch form @xmath23 where @xmath24 is the crystal volume and @xmath25 is the wave vector . inserting the wave function @xmath26 into the schroedinger equation one obtains the well - known equation for the energy @xmath27 and the coefficients @xmath28 @xmath29 where @xmath30 is the free electron mass and @xmath31 are reciprocal lattice vectors . in absence of the periodic potential there is @xmath32 , @xmath33 , and @xmath34 is the free electron energy . for a weak periodic potential we may treat @xmath18 as a perturbation and approximate @xmath28 for @xmath35 by retaining only linear terms in @xmath21 .
we obtain then @xmath36 and @xmath33 .
the perturbed energy is @xmath37 for weak potentials the correction to the free electron energy is small .
this , however , is true only if @xmath38 . for @xmath25 and @xmath39 such that @xmath40 we expect @xmath41 to be comparable to @xmath42 and the potential may not be treated as a weak perturbation .
the well known way to treat this problem is to use the approximation for nearly degenerate levels in which we neglect in ( [ eqpertb ] ) all @xmath28 except @xmath42 and @xmath41 .
we then find @xmath43 equations ( [ eq2x2 ] ) are equivalent to dealing with the hamiltonian @xmath44 which is valid for @xmath25 such that @xmath45 .
hamiltonian ( [ defhh ] ) has two eigen - energies @xmath46 and two eigen - states @xmath47 where @xmath48 @xmath49 and @xmath50 . taking only one vector @xmath51 in the reciprocal lattice we are in reality considering a one - dimensional problem . to fix our attention we consider the symmetry of a simple cubic lattice . as to the two points in the reciprocal lattice ,
see ( [ eq2x2 ] ) , it is convenient to take @xmath52 and @xmath53 $ ] , where @xmath54 is the lattice period
. then the two parabolas @xmath55 and @xmath56 in ( [ epertb ] ) cross at @xmath57 $ ] , which determines the brillouin zone boundary on the positive @xmath58 axis , see figure [ figure1 ] .
the energy gap at the zone boundary is @xmath59 .
the energies @xmath60 and @xmath61 depend in reality only on @xmath58 .
when using the nearly - degenerate perturbation theory based on ( [ defhh ] ) we should keep in mind that this procedure is only valid near the degeneracy point @xmath62 , but it progressively ceases to work as @xmath58 is lowered toward zero .
we note that @xmath61 of ( [ defed ] ) is analogous to the relativistic dispersion relation @xmath63^{1/2}$ ] .
thus @xmath64 corresponds to @xmath65 , while @xmath66 , which is quadratic in momentum , corresponds to @xmath67 .
energy _ versus _ @xmath58 for free and nearly- free electrons ( schematically ) .
the dashed line shows @xmath68 dispersion for free electrons plotted in the first brillouin zone .
the solid line indicates @xmath69 dispersion for the free motion weakly perturbed by a periodic potential.,width=321,height=321 ] it is convenient to split the hamiltonian ( [ defhh ] ) into two parts @xmath70 where @xmath71 and @xmath72 in terms of the pauli matrices the hamiltonian ( [ defhd ] ) reads @xmath73 hamiltonian @xmath74 has the form reminiscent of the dirac hamiltonian for relativistic electrons in vacuum , while the part @xmath75 is proportional to the unity matrix and it can be treated as a @xmath76-number . the decomposition ( [ decomposition ] ) is directly related to the two terms in the energy ( [ defe12 ] ) .
the quantum velocity is @xmath77 .
we calculate @xmath78 where @xmath79 and @xmath80 .
it follows from ( [ vquant ] ) that the quantum velocity @xmath81 is an operator , not a number .
since @xmath60 depends only on @xmath58 , the only non - vanishing component of @xmath82 is @xmath83 . in the following we drop the index @xmath84 .
eigen - values of the quantum velocity @xmath85 are @xmath86 .
this seems paradoxical , as it means that the quantum velocity takes only two constant ( and extreme ) values .
a similar result is obtained for the dirac equation describing relativistic electrons in vacuum , for which the eigenvalues of the quantum velocity are @xmath87 .
it is known that this feature is related to the phenomenon of zitterbewegung .
it can be easily verified that the quantum velocity ( [ vquant ] ) does not commute with the hamiltonian ( [ defhh ] ) .
this means that @xmath88 does not vanish .
the original derivation of schroedinger s is based on the quantum equation of motion ( see also @xcite ) .
let us calculate the time dependence of @xmath89 .
we have @xmath90 + [ \hat{\bm v}_{\bm k } , \hat{h}].\ ] ] since @xmath75 and @xmath91 are unity matrices , they commute with any number matrices . therefore @xmath92=0 $ ] , so that @xmath93 .
thus @xmath94 = 2 \hat{v}_{\delta}\hat{h}_{\delta } - \{\hat{v}_{\delta},\hat{h}_{\delta}\},\ ] ] where the anti - commutator of @xmath95 .
hence @xmath96 let us calculate the second time derivative of @xmath97 @xmath98 = \frac{1}{i\hbar}[2\hat{v}_{\delta}\hat{h}_{\delta } - 2{u_{\delta}}\delta,\hat{h}_{\delta}]= \nonumber \\ & = & \frac{2}{i\hbar}[\hat{v}_{\delta},\hat{h}_{\delta}]\hat{h}_{\delta } = 2\frac{d\hat{v}_{\delta}}{dt}\hat{h}_{\delta}. \end{aligned}\ ] ] this represents a differential equation for @xmath99 .
its solution is @xmath100 where @xmath101 is a constant operator . inserting ( [ vprim_t ] ) to ( [ vprim2 ] ) @xmath102 solving ( [ eqa0 ] ) for @xmath97
we obtain @xmath103 integrating ( [ vd_t ] ) with respect to time and adding the term due to @xmath84 component of @xmath104 we finally find @xmath105 in order to find @xmath101 we use ( [ eqa0 ] ) for @xmath106 @xmath107 at @xmath106 there is @xmath108 .
similarly , it follows from ( [ vd_t ] ) and ( [ defa0 ] ) that @xmath109 equals to the @xmath84 component of the initial velocity @xmath82 from ( [ vquant ] ) . in order to interpret the result ( [ z_t ] ) we observe that the eigen - energy of @xmath74 is @xmath110 , where @xmath61 is given by ( [ defed ] ) .
there is @xmath111 and @xmath112 .
the exponential term in ( [ z_t ] ) is ( see appendix ) @xmath113 the first three terms in ( [ z_t ] ) describe the classical motion .
the last term , according to ( [ exph ] ) , describes oscillations with the frequency @xmath114 .
this frequency corresponds directly to the interband energy @xmath115 , as seen in figure [ figure1 ] .
since the hamiltonian @xmath74 is a matrix , the position @xmath116 is also a matrix .
we have explicitly @xmath117 where @xmath118
the component @xmath119 has negative signs of the first two terms . further @xmath120 + \right .
\nonumber \\ \label{z21_t } & & \left
. + \frac{\delta}{{e_{\delta } } } \sin\left(\frac{2{e_{\delta } } t}{\hbar}\right ) \right\ } + \frac{{u_{\delta } } \delta v_{{\bm q}}^ * t}{{e_{\delta}}^2},\end{aligned}\ ] ] where @xmath121 and @xmath122 .
the amplitude of the oscillating term in ( [ z11_t ] ) is @xmath123 , where the zitterbewegung length is defined as @xmath124 it corresponds to the compton wavelength in relativistic quantum mechanics .
its numerical estimation is given below .
in agreement with the history of the subject , as described in the introduction , we can legitimately call the above oscillations the zitterbewegung .
we shall discuss the subject of zb more thoroughly below .
here we emphasize how _ little _ we have assumed to obtain the trembling motion - we have only perturbed the free electron motion by a periodic potential .
now we consider a more realistic picture describing an electron in terms of a wave packet .
taking , as before , @xmath125 $ ] and using the fact that @xmath60 and @xmath61 depend only on @xmath58 , we take a gaussian packet in the form @xmath126 where @xmath127 determines the packet s width and @xmath128 fixes its center in the @xmath58 space . averaging the oscillating part of the motion over the wave packet we obtain from ( [ z_th ] ) @xmath129 where @xmath130 is given in ( [ z11_t ] ) .
zitterbewegung oscillations of nearly - free electrons _ versus _ time , calculated for a very narrow wave packet centered at various @xmath128 values .
the band parameters correspond to gaas , see text.,width=321,height=321 ] amplitude of zitterbewegung of nearly - free electrons _ versus _
packet center @xmath128 , calculated for two widths of the wave packet .
symbol @xmath131 denotes the bohr radius .
material parameters correspond to gaas.,width=321,height=321 ] to begin , let us take the packet to be a delta function in the @xmath58 space centered at @xmath128 .
this corresponds to a completely non - localized packet in the real space .
we can then take various @xmath128 values beginning with @xmath132 at the zone boundary . in figure [ figure2 ]
we show the calculated zb oscillations of @xmath133 for different values of @xmath128 .
it can be seen that , as @xmath128 diminishes from @xmath134 toward the zone center , the amplitude of zb quickly drops .
this should not be surprising since , as is well known ( see figure [ figure1 ] ) , the effect of the periodic potential on the free electron motion is the strongest at the zone boundary @xmath62 , where the minimum gap occurs .
( we do not consider here the gap at @xmath135 for the upper branch ) .
figure [ figure3 ] shows the amplitude of zb for @xmath136 and @xmath137 , as calculated from ( [ pack_z11_t ] ) .
the quantity @xmath138 is the bohr radius .
for @xmath136 we deal in ( [ defpacket ] ) and ( [ pack_z11_t ] ) with a delta - function and the solid line in figure [ figure3 ] follows the dependence @xmath139 of ( [ z11_t ] ) .
when the width of the wave packet increases ( @xmath127 decreases ) the amplitude of zb for @xmath140 diminishes and , as @xmath128 is lowered , it becomes independent of the width .
since our model is not valid for @xmath128 near zero , we are limited in our considerations to not too small @xmath127 and @xmath128 values . however , since the zb amplitude diminishes so quickly with diminishing @xmath128 , it is justified to limit the considerations of zb to the vicinity of the band extremes .
it can be seen from figure [ figure2 ] that , with decreasing @xmath128 , the frequency of zb increases .
this increase follows from the behavior of the gap @xmath115 , as illustrated in figure [ figure1 ] . in order to calculate numerical values of the zb
we need to specify material parameters . as a matter of example
we take @xmath141ev and @xmath142 , corresponding to gaas .
this gives @xmath143 .
this value can be compared with @xmath144 for gaas , as obtained with the use of @xmath145 theory for the fundamental gap in gaas at @xmath146 @xcite .
the above estimation of @xmath147 based on the simple model is better than one could expect .
clearly , if we take the @xmath148 value corresponding to @xmath149ev for insb , @xmath147 would be seven times larger .
zitterbewegung contribution to the velocity for nearly - free electrons _ versus _
time , calculated for a very narrow wave packet centered at various @xmath128 values .
the band parameters correspond to gaas , see text.,width=321,height=321 ] next we calculate an observable quantity that is the electric current caused by zb .
it is given by the velocity multiplied by the charge .
the oscillatory part of velocity is given by the first term in ( [ vd_t ] ) .
we average it using the wave function ( [ defpacket ] ) which selects the component @xmath150 @xmath151 the results for the velocity , computed with the gaussian wave packet ( [ defpacket ] ) , are plotted in figure [ figure4 ] .
they are quantitatively similar to those shown in figure [ figure2 ] , the zb frequency is clearly the same , the amplitude decreases with increasing @xmath128 .
the phase , however , is different and the velocity @xmath150 is not zero at @xmath106 , but is equal to @xmath152 from ( [ vquant ] ) .
also , the velocity does not oscillate around zero , it oscillates around the value @xmath153 resulting from the second term in ( [ pack_v11_t ] ) . for @xmath132 we can find @xmath153 analytically @xmath154 \},\ ] ] where @xmath155 and @xmath156 is the error function . for narrow packets ( large @xmath127 ) the shift @xmath153 tends to zero .
as mentioned in the introduction , foldy and wouthuysen @xcite proposed a transformation which , in absence of external fields , transforms the dirac hamiltonian for relativistic electrons in vacuum into a form in which positive and negative electron energies are separated .
it was recently shown by zawadzki @xcite that similar transformations exist for the @xmath157 hamiltonians describing band structures in narrow gap semiconductors and carbon nanotubes .
since also the hamiltonian @xmath74 of ( [ defhd ] ) bears similarity to the dirac hamiltonian , we can expect that a similar transformation exists for nearly - free electrons as well .
this is indeed the case .
we define a unitary transformation @xmath158 where @xmath159 .
it is easy to verify that @xmath160 .
further @xmath161 and obviously @xmath162 since @xmath75 is proportional to the unity matrix ( see ( [ defhk ] ) ) .
thus , for the transformed hamiltonian the eigen - energy problem factorizes into two independent problems for positive and negative @xmath61 energies .
this means that the wave function corresponding to the positive energy has the lower component equal to zero while the wave function for the negative energy has the upper component equal to zero .
however , this is true also for other wave functions in the transformed representation , as we show below .
we consider an arbitrary wave function @xmath163 in the two - component representation .
it can be expressed in general in the form @xmath164 where @xmath165 let us now transform the above functions using the @xmath166 operator : @xmath167 . after some manipulations we obtain @xmath168 the @xmath169-dependent function under the square root may be put under the integral sign , as explained in @xcite . the functions @xmath170 have the above mentioned property of having only the upper or the lower non - vanishing components , which is guaranteed by the pre - factors @xmath171 .
using the inverse fourier transform @xmath172 we have @xmath173 where @xmath174 the kernels @xmath175 are not point transformations . to illustrate this we will transform the eigen - function of the position operator @xmath176 in the old representation , i.e. the dirac delta function multiplied by a unit vector @xmath177 the transformed functions are @xmath178 and @xmath179 where @xmath180 and @xmath181 it is seen explicitly that the transformed functions @xmath182 and @xmath183 have vanishing lower or upper components , respectively . both @xmath184 and @xmath185 are normalized to delta functions . to demonstrate that the transformed functions are characterized by a certain width , we calculate their second moments @xmath186 .
since @xmath187 , where @xmath188 , we have @xmath189 in ( [ defm2 ] ) we used subscripts @xmath169 and @xmath190 to indicate the variables of the integration . using relations @xmath191 and similarly for @xmath190 the triple integral is reduced to @xmath192 where @xmath147 is defined in ( [ deflambdaz ] ) . the above result has been calculated directly using @xmath60 and @xmath61 from ( [ defd ] ) and ( [ defed ] ) .
note that since @xmath193 has the dimension [ m@xmath194 , the dimension of @xmath195 is [ m ] .
similar calculations for the second moment of @xmath196 give @xmath197 . thus the transformed functions for the upper and lower energies are characterized by the same widths , as should be expected .
we will discuss physical implications of the above calculations in section vii . when transforming various wave functions from the two - component representation to the one - component representation with the use of ( [ defpsipmp ] ) , it is important to know more about the kernels @xmath198 , as given by ( [ defkpm ] ) . as an example we will calculate and plot @xmath193 given by ( [ defkp11 ] ) because both @xmath60 and @xmath61 are centered at @xmath199 , it is convenient to change the variables @xmath200 .
then we obtain @xmath201 where @xmath202 @xmath203 and @xmath204 .
the quantities @xmath205 and @xmath206 are centered at @xmath135 .
after the change of variables , we singled out the rapidly oscillating part of @xmath193 , which is related to the position of bands extremes in the @xmath25 space .
the remaining part @xmath207 is a smoothly varying function of @xmath208 with a singularity at @xmath209 .
we consider the integrand in ( [ defk11 ] ) @xmath210 for @xmath211 the function @xmath212 tends to unity as @xmath213 , while for @xmath214 it tends to zero as @xmath215 .
therefore the integral ( [ defb ] ) is poorly convergent . nevertheless , we can calculate it with the help of the heaviside function @xmath216 , which has a similar behavior to the integrand ( [ defb ] ) for @xmath217 .
we have @xmath218\exp(ik_z\zeta)dk_z + \right .
\nonumber \\ & & + \left .
\int_{-\infty}^{\infty } \theta(k_z)\exp(ik_z\zeta)dk_z \right).\end{aligned}\ ] ] the second integral is @xmath219 .
thus the real and imaginary parts of @xmath207 are @xmath220 = \frac{1}{2\pi } \int_{-\infty}^{\infty } \left [ b(k_z)- \theta(k_z)\right]\cos(k_z\zeta)dk_z + \frac{1}{2}\delta(\zeta)\ ] ] and @xmath221 = \frac{1}{2\pi } \int_{-\infty}^{\infty } \left [ b(k_z)- \theta(k_z)\right]\sin(k_z\zeta)dk_z + \frac{1}{2\pi\zeta}.\ ] ] the above integrals are carried out numerically .
the results are plotted in figure [ figure5 ] . both @xmath222 $ ] and @xmath223 $ ] are singular at @xmath209 and for large @xmath208 they decay exponentially . real and imaginary parts of the kernel of foldy - wouthuysen transformation for nearly - free electrons ( the smooth component ) _ versus _ @xmath188 .
material parameters correspond to gaas.,width=321,height=321 ] since we are dealing with weakly convergent integrals it is of interest to verify the accuracy of the above numerical calculations . to this end
, we check two sum rules holding for @xmath207 defined in ( [ defk11 ] ) ( see @xcite ) .
the first rule is @xmath224 now we verify that the same result is obtained from the numerical calculations of integrals in ( [ kzre ] ) and ( [ kzim ] ) . since the imaginary part of @xmath207 is an odd function of @xmath208 , it gives no contribution to @xmath225 .
the numerical calculation of the real part of @xmath207 gives @xmath226 with the accuracy of 10@xmath227 .
the second sum rule is related to the second moment of @xmath207 .
calculations similar to the presented above give @xmath228 taking into account the normalization ( [ defs0 ] ) we obtain the extension of @xmath229 function to be @xmath230 , which is in exact analogy to the relativistic dirac electrons @xcite , see also @xcite .
similar results are obtained for the @xmath231 and @xmath232 integrals defined using @xmath233 function .
the fw transformation is not the only transformation which can decouple positive and negative energies in the field - free case . in the relativistic quantum mechanics described by the dirac equation other transformations
were devised , see for example cini - touschek @xcite . in a recent paper mulligan
@xcite introduced a still different transformation separating the @xmath234 dirac equation into two @xmath235 equations for the electron and antielectron , respectively .
it is possible that an analogous transformation would be possible for the nearly - free non - relativistic electrons considered above .
in the preceding sections we considered the case of a weak periodic potential acting on free electrons and we showed that this potential leads to the zb . now we consider an opposite limit of strong periodic potential .
an effective treatment of a strong periodic potential is the tight - binding method .
we use here as an example the so called empirical tight binding method . in this model
one takes one @xmath236 orbital per cation and three @xmath237 orbitals per anion including nearest - neighbor and second nearest - neighbor interactions .
spherical approximation is assumed , so all @xmath238 and @xmath237 bands are isotropic and their @xmath25-dependence is given by the @xmath239 dispersion .
the model was used for calculating magnetic interactions in dilute magnetic semiconductors , approximating the band structure of cd@xmath240mn@xmath241te within the whole brillouin zone @xcite .
this scheme provides a good semi - quantitative description of both upper valence bands as well as the lowest conduction band . in the basis @xmath242 , ( c - cation , a - anion )
the hamiltonian is @xmath243 where @xmath244 , \label{deft1}\ ] ] @xmath245,\ ] ] @xmath246 + 4a_2\cos(\frac{1}{2}ak ) , \label{deft3}\ ] ] @xmath247 six parameters in ( [ deft1])-([defvk ] ) : @xmath248ev , @xmath249ev , @xmath250ev , @xmath251ev , @xmath252ev , @xmath253ev are the slater - koster parameters in notation used in @xcite , and @xmath254 is cdte lattice constant .
the parameters @xmath255 ( @xmath256 ) are cation ( anion ) on - site energies , @xmath257 is a single nearest- neighbor hopping parameter , while @xmath258 are the second - neighbor parameters .
hamiltonian ( [ defhetmb ] ) can be factorized giving one doubly degenerate energy band @xmath259 coming from @xmath260 and @xmath261 orbitals , and two energy bands coming from the @xmath262 interaction .
the @xmath263 hamiltonian can be written in the form @xmath264 where @xmath265 and @xmath266 .
the hamiltonian ( [ hsp ] ) is very similar to that for nearly - free electrons of ( [ decomposition])-([defhk ] ) .
the difference is , that @xmath267 and @xmath268 depend on the absolute value of @xmath269 .
also , now the minimum band gap occurs at @xmath270 .
all eigen - energies of the hamiltonian @xmath271 have periodicity of the lattice constant .
the quantum velocity @xmath272 does not commute with @xmath271 . to calculate @xmath273 we use the heisenberg picture @xmath274 which gives for the @xmath273 matrix @xmath275 the component @xmath276
is given by ( [ z11_etbm ] ) with the changed signs of the first two terms .
further @xmath277 - i\frac{\delta}{{e_{\delta}}}\times \right . \\
& & \left .
\sin\left(\frac{2{e_{\delta } } t}{\hbar}\right ) \right\ } -i\left(\frac{v_k^2v_k^ { ' } + v_k\delta_k\delta_k^{'}}{\hbar { e_{\delta}}^2}\right)t,\end{aligned}\ ] ] and @xmath278 . in ( @xmath279)-(@xmath280 )
the prime denotes a differentiation with respect to @xmath281 , and @xmath282 .
equations ( @xmath279)-(@xmath280 ) are formally similar to ( [ z11_t])-([z21_t ] ) , the main difference is that now there appear terms related to the dependence of @xmath268 on @xmath281 .
we calculated the zb oscillations of @xmath283 , as given by ( [ z11_etbm ] ) , using the gaussian packet of the form ( [ defpacket ] ) .
the results are similar to those illustrated in figure [ figure2 ] . in figure [ figure6 ]
we show the calculated amplitudes of zb as functions of the packet center @xmath284 for three widths of the packet . here
the @xmath284-dependence does not have a maximum at @xmath285 , because at this point the interaction @xmath268 between the bands vanishes , see ( [ defvk ] ) .
the presence of a strong periodic potential leads to two effects .
first , the quadratic dispersion relation @xmath286 for free electrons is replaced by a periodic one .
second , the potential mixes @xmath236 and @xmath287 states to form two @xmath263 energy bands .
this mixing leads to the zitterbewegung .
amplitude of zitterbewegung of tightly bound electrons _ versus _
packet center @xmath284 , calculated for three widths of the wave packet .
symbol @xmath131 denotes the bohr radius .
material parameters correspond to cdte , see text.,width=321,height=321 ]
the main result of our work is that both in the case of nearly - free electrons and in the opposite case of tightly bound electrons we predict the zitterbewegung phenomenon . comparing this result with the previous work , in which the zb was predicted with the use of lcao @xcite and @xmath9 theory @xcite ,
we conclude that zb is not due to a particular approach .
in fact , the mathematics is quite similar in all above theories and , although we deal with non- relativistic electrons , it resembles the formulation of relativistic quantum mechanics for free electrons in vacuum .
it is clear that the fundamental underlying reason for the appearance of zb in solids is the periodic potential of the lattice . particularly instructive in this respect are figures [ figure2 ] and [ figure3 ] of the present paper which show that the amplitude of zb is directly related to the effect of periodic potential on the free electron motion .
this result is not surprising . without specifying any particular band model we deal in solids with the periodic hamiltonian @xmath288 and
the electron velocity is @xmath289 .
it follows that @xmath290\neq 0 $ ] , i.e. the velocity is not constant in time because of the periodic potential @xmath18 . in this perspective the various models mentioned above simply illustrate how this result comes about . clearly , the velocity does not commute with the hamiltonian in the presence of other potentials as well
. however the periodic potential is special because , due to the bloch theorem , the electrons can propagate in the perfect crystal without scattering and the quasi - momentum @xmath291 is a good quantum number .
for this reason it is possible to treat the electrons as almost free particles and replace the influence of the periodic potential by an effective electron mass .
still , as demonstrated in refs . @xcite and the present paper , within two - band models the basic non - commutativity of @xmath292 and @xmath293 mentioned above remains in the form of non - commuting @xmath235 matrices with the resulting zitterbewegung . if the bands are completely separated , we have @xmath294 and @xmath295 , so that @xmath296 and @xmath292 commute and the zb disappears . however , there is a price to pay for this separation .
it is shown in refs .
@xcite and in the present paper that , once the electrons are described by a one - band equation ( so that their energy is completely specified ) , they should be treated as objects of a finite size .
the last effect is observable in the presence of an external potential due to appearance of the so called darwin term for free relativistic electrons @xcite as well as semiconductor electrons @xcite . as emphasized throughout our paper
( see also refs .
@xcite ) , in the two - band model the zb of non - relativistic electrons in solids is in close analogy to the zb of relativistic electrons in vacuum , as first proposed by schroedinger .
the hamiltonians for the two cases are very similar and in both systems the zb results from an interference of electron states corresponding to positive and negative electron energies @xcite .
if , with the use of foldy - wouthuysen transformation , the states of positive and negative energies are separated in the hamiltonian and in the wave functions , the zb does not occur because the positive energy state ( say ) has nothing to interfere with .
this corresponds to the separation of bands mentioned above and the conclusions of the two reasonings agree .
thus we are confronted with the following choice : 1 ) we use a two- band description , the electrons are point - like particles and they experience the zitterbewegung .
2 ) we use a one - band description , the electrons do not experience the zitterbewegung but they are characterized by a quantum radius of the size equal to the zb amplitude . the last point is illustrated by ( [ defs2 ] ) describing the average smearing of the transformed delta function .
it is equal to the amplitude of zb given by ( [ z11_t ] ) .
one can say that the separation of energy bands by the foldy - wouthuysen transformation is equivalent to a certain averaging of the zb motion .
it was observed @xcite that , since the zitterbewegung had been predicted for plane dirac waves , it is not quite clear what the trembling motion means for an electron uniformly distributed in space . in this connection
it is important that the amplitude does not vary much when the electron is represented by a wave packet localized in real space , see figure [ figure3 ] .
passing to more specific points of our treatment , we emphasize again how little we had to assume to derive the zb in case of nearly - free electrons - it was enough to perturb the free electron motion by a periodic potential .
this case has certain particularities . in the typical two - band situations , both in vacuum @xcite and in solids @xcite
, there exists a maximum velocity in the system which plays an important role in the theory . in case of nearly - free electrons
there is no maximum velocity since the perturbed energy branches tend asymptotically to the free electron parabola @xmath297 , see figure [ figure1 ] .
this is reflected in the velocity @xmath298 , while the velocity @xmath97 has the typical two - band behavior and it is responsible for the zb .
in fact , the velocity @xmath97 has the maximum value .
it is equal to @xmath299 , see our eigen - value considerations after ( [ vquant ] ) . for @xmath142
the maximum velocity is 6.4@xmath300cm / s , which should be compared with @xmath301cm / s obtained from the @xmath9 theory for gaas and other iii - v compounds @xcite .
again , our simple model gives quite a reasonable estimation .
in the nearly - free electrons model @xmath147 is proportional to @xmath302 ( see ( [ deflambdaz ] ) ) , which agrees with the @xmath9 approach @xcite , where @xmath303 . in narrow gap materials @xmath147
can be as large as tens of angstroms .
as far as the phenomenon of zb is concerned , the behavior of nearly - free and tightly bound electrons is quite similar .
the main difference comes from the fact that in nearly - free case the fourier coefficients of the periodic potential do not depend on the wave vector @xmath25 , whereas in the tightly bound case they go to zero for vanishing @xmath281 . as a result , in the first case ( see figure [ figure3 ] ) the zb amplitude is highest for the @xmath128 corresponding to the minimum energy gap , while in the second case ( see figure [ figure6 ] ) the maximum amplitude is shifted with respect to the minimum gap . for the tightly bound case the electrons in @xmath304 bands ( see ( [ deft3 ] ) ) would not exhibit the zb , but we do not insist on this point since these two bands are not realistically described by the model . in order to illustrate that , after the foldy - wouthuysen transformation is carried out , the former dirac delta function is smeared into the kernel @xmath305 of ( [ defkp11 ] ) , we not only calculate it numerically ( see figure [ figure5 ] ) , but also calculate its second moment @xmath306 ( see ( [ wynm2 ] ) ) and the sum rule @xmath307 of ( [ defs2 ] ) .
the second moment has the advantage of using the standard quantum mechanical probability distribution @xmath308 , see ( [ defm2 ] ) .
its disadvantage is that this probability distribution is normalized to the dirac delta function and not to a number . as to the sum rules using the single @xmath229 function
, their advantage is that both the normalization ( [ defs0 ] ) and the sum rule @xmath309 ( [ defs2 ] ) are numbers , and ( [ defs2 ] ) can be simply interpreted as a square of smearing ( this procedure was used for free relativistic electrons by rose @xcite ) .
the disadvantage of using the single @xmath229 is that we have to separate out its smooth part , see ( [ defsmooth ] ) .
the oscillatory part @xmath310 would not appear if we considered the nearly - free electron gap at @xmath135 . methods to observe the zitterbewegung should clearly be adjusted to the investigated materials , but it seems that an appropriate tool would be the scanning probe microscopy which can produce images of coherent electron flow @xcite .
this technique uses a sharp mobile tip which can sense the electron charge .
if one used dilute magnetic semiconductors of cdmnte type @xcite , one could employ magnetic effects caused by electron oscillations .
the second category of possible observable effects is related to the problem of what happens when electrons are confined to dimensions smaller than @xmath147 . in relativistic quantum mechanics
one finds statements that a measurement of the position of a particle , if carried out with greater precision than the compton wavelength , would lead to pair production @xcite .
it is clear , however , that the pairs created this way can only be virtual , otherwise their recombination would lead to the production of energy out of nothing . the virtual carriers could be observed in screening or in magnetism .
an alternative point of view states that a stiff confinement is equivalent to an infinite potential well having the width @xmath311 , and the electrons will simply occupy the lowest energy level in such a well .
it is , however , certain that if an electron is confined to dimensions @xmath311 , its energy ( or its uncertainty ) is of the order of the gap @xmath1 between the positive and negative electron energies , which means that the one - band description is not adequate .
we are then back to the two - band model , which was our starting point .
there remain many unanswered questions concerning the trembling motion but it appears that in crystalline solids it represents a rule rather than an exception . according to the theory , the zb in semiconductors has decisive advantages over the corresponding effect in vacuum .
thus an experimental detection of the trembling motion in solids appears to be a matter of near future .
we considered theoretically non - relativistic nearly - free electrons in solids for which the periodic potential of the lattice may be treated as a weak perturbation on the free electron motion . using the two - band model we showed that electrons experience the trembling motion ( zitterbewegung ) in absence of external fields , similar to that for free relativistic electrons in vacuum .
the frequency of zb and its amplitude were derived .
the frequency is @xmath0 where @xmath1 is the energy gap between the two bands .
the amplitude @xmath147 depends on the strength of periodic potential and the lattice period . for typical parameters
@xmath147 can be of the order of 10 to 100 that is @xmath312 times larger than in vacuum .
the trembling motion is also considered for nearly - free electrons represented by wave packets , it is shown that the amplitude is not strongly dependent on packet s width .
the foldy - wouthuysen type of unitary transformation , known from relativistic quantum mechanics , is used to separate the energy bands .
consequences of the fw transformation are investigated .
it is demonstrated that , if one uses a one - band description , the electrons do not experience the trembling motion but they should be treated as particles having size @xmath147 . tightly bound electrons are considered as well to provide an opposite case to nearly - free electrons . within the two - band model the trembling motion is obtained also in this case demonstrating that the zb phenomenon is not related to a specific theoretical approach .
it is concluded that the trembling motion is directly related to the effect of the periodic potential on the electron and , as such , it should occur in many situations in solids .
this work was supported in part by the polish ministry of sciences , grant no pbz - min-008/p03/2003 .
[ [ section ] ] first , we prove some properties of @xmath74 of ( [ defhd ] ) .
the eigen - values of @xmath74 are @xmath110 .
if @xmath313 and @xmath314 are eigen - states of @xmath74 , then @xmath315 are two projection operators .
it follows that @xmath316 and @xmath317 .
accordingly @xmath318 since @xmath319 , we have @xmath320
. then @xmath321 because @xmath322 . for a real @xmath323
there is @xmath324 because @xmath325 we have @xmath326\hat{p}_1 + \nonumber \\ & + & [ \cos({e_{\delta } } b ) - i\sin({e_{\delta } } b)]\hat{p}_2.\end{aligned}\ ] ] grouping the terms with cosine and sine functions we get @xmath327 this identity is used in ( [ exph ] ) .
99 schroedinger e 1930 _ sitzungsber .
preuss . akad .
phys . math .
kl . _ * 24 * 418 ; schroedinger s derivation is reproduced in @xcite rose m e 1961 _ relativistic electron theory _
( new york : wiley ) bjorken j d and drell s d 1964 _ relativistic quantum mechanics _
( new york : mcgraw - hill ) greiner w 1994 _ relativistic quantum mechanics _ ( berlin : springer ) foldy l l and wouthuysen s a , 1950 _ phys .
rev . _ * 78 * 29 pryce m h l 1948 _ proc .
. london _ * a195 * 62 tani s 1951 _ progr .
* 6 * 267 newton t d and wigner p 1949 _ rev
* 21 * 400 huang k 1952 _ am .
_ * * 20 479 feshbach h and villars f 1958 _ rev .
phys . _ * 30 * 24 lock j a 1979 _ am .
_ * 47 * 797 barut a o and bracken a j 1981 _ phys .
_ d * 23 * 2454 barut a o and malin s 1968 _ rev .
_ * 40 * 632 krekora p , su q and grobe r 2004 _ phys . rev . lett._*93 * 043004 cannata f , ferrari l and russo g 1990 _ solid state commun . _ * 74 * 309 ; ferrari l and russo g 1990 _ phys . rev .
_ b * 42 * 7454 zawadzki w 2005 _ phys .
_ b * 72 * 085217 zawadzki w 2006 _ phys .
_ b * 74 * 205439 zawadzki w 1970 in _ optical properties of solids _ ed haidemenakis e d ( new york : gordon and breach ) p 179 zawadzki w 1997 in _ high magnetic fields in the physics of semiconductors ii _
ed landwehr g and ossau w ( singapore : world scientific ) p 755 schliemann j , loss d and westervelt r m 2005 _ phys .
* 94 * 206801 ; 2006 _ phys .
_ b * 73 * 085323 katsnelson m i 2005 _ europ .
phys . journal _
b * 51 * 157 jiang z f , li r d , zhang s c and liu w m 2005 _ phys .
_ b * 72 * 045201 cserti j and david g 2006 _ phys . rev .
_ b * 74 * 172305 cini m and touschek b 1958 _ nuovo cimento _ * 7 * 422 mulligan b _ annals of physics _ 2006 * 321 * 1865 larson b e , hass k c , ehrenreich h and carlsson a e 1998 _ phys .
_ b * 37 * 4137 topinka m a , leroy b j , shaw s e j , heller e j , westervelt r m , maranowski k d and gossard a c 2000 _ science _ * 289 * 2323 leroy b j 2003 _ j. phys . :
condens . matter _
* 15 * r1835 | we show theoretically that nonrelativistic nearly - free electrons in solids should experience a trembling motion ( zitterbewegung , zb ) in absence of external fields , similarly to relativistic electrons in vacuum .
the zitterbewegung is directly related to the influence of periodic potential on the free electron motion .
the frequency of zb is @xmath0 , where @xmath1 is the energy gap .
the amplitude of zb is determined by the strength of periodic potential and the lattice period and it can be of the order of nanometers .
we show that the amplitude of zb does not depend much on the width of the wave packet representing an electron in real space .
an analogue of the foldy - wouthuysen transformation , known from relativistic quantum mechanics , is introduced in order to decouple electron states in various bands .
we demonstrate that , after the bands are decoupled , electrons should be treated as particles of a finite size .
in contrast to nearly - free electrons we consider a two - band model of tightly bound electrons .
we show that also in this case the electrons should experience the trembling motion .
it is concluded that the phenomenon of zitterbewegung of electrons in crystalline solids is a rule rather than an exception . |
global evolution of protoplanetary disks ( ppds ) is governed by the processes of angular momentum transport and outflow mass loss .
these processes directly control disk structure and evolution , which set the timescales of disk dispersal and hence planet formation ( e.g. , see @xcite and @xcite for reviews )
. they also strongly affect the evolution of dust grains , which are building blocks of planets , and feedback to disk thermal and chemical structures ( e.g. , see @xcite and @xcite for reviews ) .
if planets are formed within the disk lifetime , planet - disk interaction leads to planet migration , which is also sensitive to global disk structure and evolution ( e.g. , see @xcite for a review ) . in brief
, a reliable global evolutionary picture of ppd is key to understanding most processes of planet formation .
most ppd models used in planet formation research are constructed upon the viscous @xmath0 disk model @xcite , where the underlying assumption is that the disk is turbulent presumably due to the magneto - rotational instability ( mri , @xcite ) .
however , the extremely weak level of ionization introduces strong non - ideal magnetohydrodynamic ( mhd ) effects that suppress or damp the mri in most regions in ppds , while pure hydrodynamic mechanisms appear unable to provide sufficiently level of turbulent viscosity ( e.g. , @xcite ) .
angular momentum transport in ppds is thus most likely governed by magnetized disk winds , as demonstrated in disk simulations that properly taking into account these non - ideal mhd effects ( e.g. , @xcite ) . in the mean time
, the mri may operate in the surface layer of the outer disk , owing to strong far - uv ( fuv ) ionization at disk surface @xcite , which can lead to vigorous turbulence and mediate a certain level of viscous transport of angular momentum @xcite .
we note that in the conventional studies of mhd winds , wind launching generally requires near equipartition field at the disk midplane ( e.g. , @xcite ) . as a result
, the disk surface density must be very low to be consistent the observed disk accretion rates ( e.g. , @xcite , otherwise , accretion rate would become too high ) , making the disk wind scenario less appealing to account for the mass content of gas and dust in ppds .
on the other hand , in the aforementioned more realistic simulations , winds are launched from several scale heights above midplane , because the midplane region is too weakly ionized for sufficient coupling between the gas and magnetic fields .
with much weaker field ( magnetic pressure much less than midplane gas pressure ) permitted for wind launching , the new scenario simultaneously accounts for the accretion rates and mass budget from observations .
the paradigm shift towards wind - driven ppd evolution calls for a model framework in replacement of @xmath0-disk models .
the aforementioned simulations are all local in vertical extent , and hence are unable to provide reliable estimates of wind kinematics .
an initial study by @xcite took the fitting relations of viscous stress and wind torque from @xcite s outer disk simulations and found slow disk evolution followed by rapid dispersal .
disk mass loss was not included in their study .
a more reliable disk evolution framework would require better determination of the wind torque and wind mass loss rate , and their dependence on physical parameters .
recently , @xcite ( hereafter b16 ) proposed a physically motivated , semi - analytical 1d model of ppd disk winds with global treatment of wind kinematics .
the model further takes into account thermodynamical effects and unifies the classical ( cold ) mhd disk wind with photoevaporation . despite remaining uncertainties explicitly discussed there , it now becomes possible to incorporate the key results into a framework of global wind - driven ppd evolution , and explore in parallel the role played by magnetic fields and thermodynamics
this is the goal of this paper .
we describe our formalism and methodology in section [ sec : model ] . in section [ sec : prof ] , we discuss the general angular momentum transport and mass loss processes without evolving the disk , and study parameter dependence .
disk evolutionary properties are studied in section [ sec : evolve ] .
we extend our model to discuss additional effects including level of disk flaring and x - ray heating in section [ sec : ext ] .
implications and limitations of our study are discussed in section [ sec : disc ] .
we summarize and conclude in section [ sec : sum ] .
we construct a 1 d disk model on the evolution of the disk surface density @xmath1 as a function of disk cylindrical radius @xmath2 in the presence of turbulence , wind torque and mass loss ( section [ ssec : eqs ] ) . in the mean time
, we adopt a simple treatment of disk vertical structure to estimate the vertical location @xmath3 where external far - uv ( fuv ) radiation can penetrate , and @xmath4 ( @xmath5 ) where the wind is launched ( wind base ) , based on which we calculate the transport properties ( section [ ssec : model ] ) .
hence , our model can be considered as 1 + 1 d. we discuss calculation procedures and model parameters in section [ ssec : proc ] .
we begin by writing down the equations governing the surface density evolution , incorporating the effects of viscosity , wind torque and mass loss .
let @xmath6 be the net accretion rate at cylindrical radius @xmath2 .
we adopt the sign convention such that @xmath7 is positive for net accretion .
let @xmath8 be the cumulative wind mass loss rate enclosed within radius @xmath2 .
we will always use its differential form @xmath9 where @xmath10 is the radius of the disk inner edge beyond which the wind is launched .
the bulk of ppds is cold , and unless unrealistically strongly magnetized , rotation is largely keplerian , with specific angular momentum @xmath11 , where @xmath12 , @xmath13 are the keplerian speed and corresponding angular frequency .
let @xmath14 be the vertically integrated total radial angular momentum flux .
it is given by @xmath15 where @xmath16 accounts for angular momentum flux generated from internal stresses , including reynolds stress , maxwell stress , and potentially stress from disk self - gravity ( not included above ) in the case of a massive disk that is gravitationally unstable , with overline representing averaging over the local volume .
here , we follow the @xmath0 convention , with @xmath17 being disk sound speed . note that @xmath0 can be position dependent , and we use @xmath18 to represent an effective , ( density - weighted ) vertically - averaged value .
let @xmath19 be the total vertical angular momentum flux extracted from the disk ( by the mhd wind ) enclosed within radius @xmath2 .
we only use its differential form : @xmath20 where @xmath21 and @xmath22 is called the wind lever arm ( dimensionless ) . for wind launched from radius @xmath23 ,
the physical meaning of @xmath24 is that it is the ratio of specific angular momenta in the wind flow and in the keplerian disk at @xmath25 , and it is related to wind alfvn radius @xmath26 by @xmath27 .
the @xmath28 term represents the _ excess _ angular momentum extracted from the disk , related to the wind torque . in the above
, we have dropped the hydrodynamic term in the definition of @xmath29 , which corresponds to the standard definition of the wind base @xmath4 @xcite : at the wind base , @xmath30 .
disk continuity and angular momentum conservation equations read @xmath31 @xmath32 with preparations above , they can be combined to @xmath33-\frac{{\partial}\dot{m}_{\rm loss}}{{\partial}r}\ , \ ] ] where the two terms in the parenthesis represent accretion rate driven by viscosity and wind , respectively , and the equation can further be rearranged to @xmath34\\ & + \frac{{\partial}}{{\partial}r}\bigg[2(\lambda-1)r \bigg(\frac{{\partial}\dot{m}_{\rm loss}}{{\partial}r}\bigg)\bigg]-\frac{{\partial}\dot{m}_{\rm loss}}{{\partial}r}\ .
\end{split}\label{eq : sigevl}\ ] ] this is the master equation that we will solve to study global disk evolution . is a factor @xmath35 of that in a viscous evolution equation .
this is because we directly parameterize @xmath0 from the @xmath36 component of the stress tensor @xmath37 ( the terms in the parenthesis in equation ( [ eq : w ] ) ) , whereas assuming navier - stokes viscosity @xmath38 , @xmath37 involves @xmath39 which yields an extra factor of @xmath40 . ]
it clearly generalizes the viscous evolution equation to include wind - driven accretion and mass loss terms ( the 2nd and 3rd terms on the right hand side ) . to evaluate these terms , we must determine @xmath41 , @xmath22 and @xmath42 at each radius , as we do in the next subsection .
we follow and extend the work of b16 and construct the vertical dimension of our wind model to calculate wind properties , and the basic picture is illustrated in figure [ fig : cartoon ] .
we assume that the disk follows a two - temperature profile with @xmath43 in the disk interior and @xmath44 at disk surface ( atmosphere ) , and transitions at vertical height @xmath45 .
both layers are treated as vertically isothermal .
disk temperature @xmath46 is largely determined by stellar irradiation .
we use a power - law temperature profile similar to the standard minimum - mass solar nebular ( mmsn ) model @xcite with @xmath47 where @xmath48 is radius measured in au , and @xmath49 is the power law index . by default , we take @xmath50 , as in the mmsn model . the corresponding disk isothermal sound speed @xmath51 and disk scale height @xmath52 satisfy @xmath53 . with @xmath54 increasing with @xmath2 ,
the disk is flared , as is well known from sed modeling of t tauri disks @xcite .
disk surface ( atmosphere ) temperature is largely determined strong heating from stellar far - uv ( fuv ) radiation and x - rays . for simplicity , we assume it is a constant factor higher than disk temperature @xmath55 and hence atmosphere sound speed @xmath56 . in practice , we take @xmath57 in our calculations , consistent with typical results from more detailed thermo - chemical calculation ( e.g. , @xcite ) , as well as observational constraints from the hd 163296 disk @xcite . in the mean time , fuv is capable of fully ionizing carbon and sulfur species , leading to an ionization fraction @xmath58 , and gas and magnetic fields become well coupled in the fuv layer @xcite .
we note that x - rays alone generally produces much lower level of ionization in the surface layer ( e.g. , cf .
figure 1 of @xcite ) . launching of mhd winds
requires sufficient coupling between gas and magnetic fields , and hence it mainly relies upon fuv ionization .
let @xmath59 be the characteristic penetration depth of fuv photons .
its value is uncertain and depends on the fuv luminosity from the protostar , as well as the abundance of very small grains .
here we adopt the calculation results by @xcite , quoting @xmath60 and treat it as a constant at all radii .
the location of the fuv front is determined by tracing radial rays from the central star , until the column density traversed by the rays equals to @xmath59 ( see section [ sssec : base ] for details ) . by default , we assume that the penetration depth of fuv and x - rays are comparable so that @xmath3 is determined by @xmath59 .
in other words , we assume the rapid increase of temperature at @xmath45 is accompanied by a transition into ideal mhd regime at disk surface . in reality , x - rays may be able to provide heating into deeper layers .
this will be discussed in section [ ssec : xray ] , where we relax this assumption . , and we use @xmath4 to denote the wind base where the wind is launched . in general , @xmath61 unless net vertical magnetic field is not sufficiently strong . in this case , the warm surface layer can become mri active ( achieved at the outer disk ) , and @xmath62 , sandwiching a warm mri layer . not reflected in this cartoon picture is that the disk is flared , making fuv radiation easier to penetrate into the outer disk.,width=340 ] another important location is the wind base @xmath4 , from which the wind is launched .
the wind model of b16 assumed @xmath63 motivated by numerical simulations of the inner disk @xcite . towards the outer disk , however , fuv can effectively penetrate deeper into the disk ( in terms of @xmath64 ) as the disk is flared , and this can make the surface layer subject to the mri @xcite .
a disk wind can also be launched from the surface mri turbulent layer in the presence of net vertical field @xcite , and local simulations applicable to the outer ppds found evidence that disk outflow is launched from a higher position than the fuv front @xcite .
we describe a prescription in section [ sssec : base ] to approximately determine the location of @xmath62 in this case , and treat the layer @xmath65 as being mri active .
we apply the b16 model in the wind zone at @xmath66 , from which we estimate the wind mass loss rate and wind torque .
although the b16 model was constructed assuming laminar flows and does not strictly apply to estimate the wind kinematics of outflows launched from a turbulent surface , we assume that it can still provide a reasonable approximation in terms of wind mass loss rate and angular momentum transport .
below , we first construct the wind model assuming the locations of @xmath3 and @xmath4 are known , then we iteratively determine their locations .
the b16 wind model is constructed from the wind base @xmath4 .
being a wind model that takes into account both magnetic and thermal effects , the most important parameters are the net vertical magnetic field strength @xmath67 threading the disk and wind / atmosphere temperature @xmath68 ( or the atmosphere sound speed @xmath69 ) . we treat them as our main physical parameters , and they correspond to @xmath70 and @xmath71 defined in the b16 model .
we will discuss our choice of @xmath67 in section [ sssec : bz ] .
concerning wind temperature @xmath68 , characterized by the @xmath72 factor in ( [ eq : f ] ) , we apply an additional constraint such that the ratio of surface sound speed to keplerian velocity @xmath73 .
the b16 model tends to be less reliable at higher temperature because the slow magnetosonic point is generally located below the wind base , violating the assumptions made there .
however , we expect this compromise to have only minor impact in our calculations . as discussed in b16 ( see their figure 12 ) ,
wind kinematics is not very sensitive to wind temperature , which we also confirm in our global disk calculations in section [ sssec : tatm ] .
we fix other parameters considered in b16 at their fiducial values , where it was found that the wind solutions are generally not sensitive to these parameters .
in particular , wind properties are not very sensitive to the field inclination angle @xmath74 because the wind is warm .
we thus choose representative field geometry parameters @xmath75 .
we also choose the field divergence parameter to be @xmath76 .
scaling ( see equation 2 of b16 for definition ) , and the wind solutions do show relatively strong dependence on @xmath77 .
we choose the fiducial number @xmath76 to avoid more extreme cases ( i.e. , diverging from launching point with @xmath78 and field lines being parallel to large distances ) .
] we refer to section 2 of b16 for further model details . at each radius @xmath2 ,
the disk zone at @xmath79 satisfies hydrostatic equilibrium , and gas density profile @xmath80 within the disk is simply @xmath81}\quad ( z < z_b)\ ] ] where @xmath82 is midplane gas density . at the location @xmath45
, pressure balance requires a density jump between the disk side and the atmosphere side @xmath83 and we define @xmath84 .
while this jump should occur more smoothly in reality where temperature transition is more smooth , the resulting uncertainty can be absorbed into @xmath59 . between @xmath85 , another hydrostatic equilibrium
is maintained as @xmath86}\quad ( z_a\leq z\leq z_b)\ , \label{eq : rhomri}\ ] ] where @xmath87 is the scale height of the disk atmosphere .
typically , the wind base is located at several disk scale heights above midplane , and regions with @xmath88 encloses most of the disk mass .
thus , midplane density @xmath82 can be determined from surface density @xmath89 once @xmath3 is known . for the purpose of locating the wind base
( see section [ sssec : base ] ) , we also need the density profile in the vicinity of the wind base so as to calculate the column densities for fuv ray tracing . due to magnetic forces
, it can deviate substantially from hydrostatic equilibrium . from figure 3 of b16
, we see that at fixed wind temperature @xmath90 , the wind density profile is almost independent of wind magnetization up to a substantial distance ( @xmath91 times wind launching radius ) from the wind base .
we have also verified this trend with other choices of wind temperature .
for wind launched from radius @xmath23 , we find a reasonable fitting formula for gas density profile along the wind streamline @xmath92}\ , \label{eq : rhowind}\ ] ] where @xmath93 is gas density at the wind base from ( [ eq : rhomri ] ) , and @xmath2 is cylindrical radius along the wind streamline .
this formula is accurate to within order unity ( most cases within @xmath94 ) up to @xmath95 for wind temperature range @xmath96 .
because this density profile is solely used for finding the wind base location , we may convert it into a vertical density profile at radius @xmath23 and @xmath66 for convenience @xmath97}\ , \label{eq : rhowind1}\ ] ] because fuv photons penetrate through approximately the same column density in these two cases to reach the next wind streamline .
net vertical field strength threading the disk @xmath98 is associated with the poloidal field at the wind base @xmath70 by @xmath99 , where we fix inclination angle @xmath75 .
it is then reflected in the wind base alfvn speed @xmath100 , a crucial parameter in the b16 wind model . the radial profile of @xmath101 , and its time evolution , however , are largely unconstrained .
attempts to study magnetic flux transport so far have not considered effects other than viscosity and ohmic resistivity , nor on the effect of wind - driven accretion @xcite . at disk surface .
we expect ppd wind to lie in the opposite limit .
] we here simply consider a phenomenological approach , as adopted in @xcite , and assume that magnetic flux is distributed in such a way that midplane plasma @xmath102 of the net vertical field , @xmath103 is constant , where @xmath104 is midplane gas pressure .
we further consider two scenarios on the evolution of total magnetic flux @xmath105 . in the first scenario , we assume @xmath106 is conserved during disk evolution .
therefore , as accretion proceeds , disk becomes more strongly magnetized . in the second scenario
, we assume that the disk loses magnetic flux in a way such that @xmath106 is proportional to total disk mass @xmath107 .
this treatment avoids significant accumulation of magnetic flux and results in slower evolution .
it is conceivable that magnetic flux evolution can have major impact on global disk structure and evolution , and our phenomenological treatment bares significant uncertainties . for instance , it has been found that the mri can concentrate vertical magnetic flux into ring - like structures @xcite , accompanied by radial pressure variations known as zonal flows @xcite .
hl tau like structures @xcite may simply be a manifestation of such zonal flows . before a more reliable theory of magnetic flux transport is available
, we mainly discuss the general features of wind - driven disk evolution without focusing on the details in the evolution of surface density profiles , and explore the general role played by macroscopic parameters such as total magnetic flux and fuv penetration depth so as to better understand the interplay between magnetic and thermal effects on disk evolution and dispersal .
once the vertical density profile is obtained from our 1 + 1d calculation , we trace radial rays from the origin ( stellar location ) with different inclination angles @xmath74 ( w.r.t .
disk midplane ) to calculate the column density @xmath108 traversed by these rays .
the location of the fuv ionization front @xmath3 is determined by @xmath109 .
while this simple ray - tracing procedure ignores important effects of scattering and self - shielding @xcite , particularly on the lyman-@xmath0 photons , which dominates fuv luminosities @xcite , we again expect the uncertainties can be absorbed into the value of @xmath59 , as a first approximation . additionally , knowing the net vertical field strength @xmath98
, we can estimate the plasma @xmath102 of the net vertical field at @xmath45 : @xmath110 .
we note that in the ideal mhd regime , the most unstable mri wavelength is given by ( e.g. , @xcite ) @xmath111 due to the rapid density density drop with height , we expect the mri to be able to effectively operate when @xmath112 . in practice , we assume that the disk surface becomes mri turbulent when @xmath113 . otherwise , we directly set @xmath63 .
if @xmath113 , we would expect the wind to be launched from higher location from the mri turbulent layer .
we further follow the density profile using ( [ eq : rhomri ] ) and determine the wind base location @xmath4 by setting @xmath114 .
this treatment is by no means rigorous , yet without better knowledge of the wind launching process from turbulent disk surfaces and its global kinematics , we expect it to be a reasonable first approximation .
local simulations of @xcite lend support to this treatment in that following a generalized ( yet unproved ) procedure to determine @xmath4 , they found that @xmath4 is located above the fuv ionization front , and becomes higher when net vertical field is weaker .
we also note that in their simulations where the disk becomes largely laminar ( ad30au1e3 and ad30au1e4l ) , @xmath115 at the wind base is about @xmath116 and @xmath117 , which are in rough agreement with our choice of threshold @xmath118 .
the procedures above to determine @xmath3 and @xmath4 can be iterated with the vertical density profile calculations described in section [ sssec : den ] .
note that iteration only needs to be done at the beginning to establish the initial condition , where convergence can be achieved to within @xmath119 typically within 3 - 4 iterations .
the subsequent disk evolution is sufficiently slow that no iteration is necessary .
pure hydrodynamic instabilities , such as the vertical shear instability @xcite , convective overstability and baroclinic vortex amplification@xcite , and the zombie vortex instability @xcite , may operate in certain regions of ppds largely depending on disk thermodynamics . in general , these instabilities are found to produce very limited viscous transport with @xmath120 . in this work , we adopt @xmath121 at locations @xmath79 to represent residual viscosity " from such hydrodynamic instabilities . as we shall see , because of the dominant role played by disk winds , the exact value of @xmath122 adopted here is unimportant . if an mri active zone is present , we set @xmath123 in this region ( @xmath65 ) .
we choose a relatively large @xmath0 because the net vertical field at this surface layer is effectively strong ( with @xmath115 of the order 100 ) , and it is well known from mri simulations ( e.g. , @xcite ) that the resulting @xmath0 is larger than zero net flux case and is of the order 0.1 or higher . additionally , while transport by the mri is suppressed or strongly damped by non - ideal mhd effects at @xmath79 , there is still maxwell stress resulting from either weak mri turbulence or large - scale fields in the outer disk , and its strength increases with net vertical magnetic flux . incorporating all these considerations
, we adopt the following form of @xmath0 in the disk zone at @xmath79 : @xmath124\ , \ ] ] where the dependence on midplane @xmath115 approximately reflects the stress level found from local simulations of layered accretion in the presence of net vertical field ( e.g. , @xcite ) . on @xmath115 is shallower in fully mri turbulent disks @xcite . ] this way , the @xmath125 value ranges from @xmath122 for the weak field case to @xmath126 for the strong field case which joins the @xmath0 value in the surface mri active zone .
although this prescription is motivated by the outer ppd gas dynamics , we simply apply it to the entire disk , allowing @xmath0 to increase with net @xmath98 . towards the inner disk where the hall effect is important , this treatment also roughly reflects the fact that the large - scale stress ( magnetic braking ) resulting from the hall - shear instability increases with net vertical flux @xcite .
we further comment that the overall disk evolution is insensitive to the exact prescription of @xmath0 values because , as we will see , disk evolution is largely wind - driven .
we apply the @xmath0 values in the disk and surface regions discussed above to equation ( [ eq : w ] ) , from which an effective value , @xmath18 , can be determined .
note that we quote the @xmath18 value based on sound speed in the disk interior @xmath51 . knowing the poloidal alfvn speed and sound speed @xmath127 and @xmath69 at the wind base
, we fit the numerical results presented in b16 to evaluate @xmath22 and @xmath42 as needed to evolve the master equation ( [ eq : sigevl ] ) .
the wind mass loss rate is characterized by the dimensionless mass loading parameter @xmath128 given by ( see equation 21 of b16 ) @xmath129 the wind is considered to be heavily loaded when @xmath130 .
figure 5 of b16 shows the dependence of @xmath128 on @xmath127 for three different wind temperatures , which we find can be well fitted by @xmath131 where @xmath132 .
we have also examined additional wind solutions and confirm that this fitting formula is accurate as long as @xmath133 ( note that it becomes unphysical if @xmath134 ) . for all models considered in this work ,
the disk surface is sufficiently warm and satisfies this condition .
the magnetic lever arm @xmath22 is defined as @xmath135 , where @xmath23 is the wind launching radius and @xmath26 is the alfvn radius . as shown in figure 7 of b16 , @xmath136 follows a well defined relation with @xmath128 .
the relation is independent of wind temperature when the wind is lightly loaded ( @xmath137 ) , and is largely parallel to the corresponding relation in the @xcite wind . with these considerations ,
we obtain a reasonable fitting relation @xmath138}}{[1+(0.2\mu)^{1/4}(c_{s , a}/v_k)^2]^2}\ ] ] we find this fitting relation is accurate for @xmath128 up to 100 for the range of @xmath139 considered in this work ( @xmath140 ) .
note that this fitting relation becomes invalid and would predict @xmath141 for very large @xmath128 and when the wind is warm . in reality , the wind would transition into a pure thermally driven towards higher temperature , and no longer extracts disk angular momentum .
note that the results presented in b16 all have @xmath142 , partly because a wind that is too heavily loaded generally violates the assumptions made there ( e.g. , the slow magnetosonic point is located below the wind base ) . in practice , because we have imposed the condition @xmath143 at the wind base ( relatively strong field ) , we never encounter a situation with @xmath144 . as a fiducial example and initial condition for our calculations
, we consider a disk model with a power - law radial profile and an exponential cutoff motivated from sub - millimeter observations @xcite @xmath145 where @xmath48 is radius measured in au .
we choose the cutoff radius to be @xmath146 au throughout this work .
the total disk mass is about 0.035 @xmath147 .
our calculations are carried out on a logarithmic grid with inner boundary at @xmath148 au , and outer boundary at @xmath149 au using 200 grid points .
although the disk - magnetosphere boundary likely lies further in , we choose @xmath148 au because the innermost disk region ( @xmath150 au ) is sufficiently hot for thermal ionization of alkali species and is expected to be fully turbulent due to the mri ( e.g. , @xcite ) .
the dynamics in this region is complex and is very sensitive to disk thermodynamics ( e.g. , @xcite ) , yet a comprehensive picture has not been established ( especially with the addition of mhd disk winds ) .
we do not attempt to model the dynamics of this region , but we also point out that this region contains only a very small fraction of total disk mass and should not affect the bulk of disk evolution . the viscous term in equation ( [ eq : sigevl ] ) is integrated with standard zero torque boundary conditions . the wind - driven accretion ( advection ) term is integrated using the standard upwind method .
we set @xmath151 g @xmath152 as a floor of surface density .
our model mainly consists of three sets of parameters .
first , magnetic flux distribution and evolution .
we consider three different values initial magnetic field strength , corresponding to plasma @xmath102 of the net vertical field at midplane @xmath153 , @xmath154 , @xmath155 and @xmath156 , with @xmath157 as fiducial value . in each case , we evolve the magnetic flux either assuming flux conservation @xmath158 const , or assuming @xmath159 .
second , fuv penetration depth .
we consider @xmath160 g @xmath161 and 0.1 g @xmath161 , taking the former as fiducial .
third , disk atmosphere temperature .
we consider two values of @xmath162 or @xmath163 , with the former as fiducial ( see equation ( [ eq : f ] ) ) .
we will discuss in section [ sec : ext ] further extensions of our model , discussing the effect of x - ray heating and the level of disk flaring .
we begin by discussing the general features without evolving the disk .
we draw the disk surface density profile from ( [ eq : disk ] ) and follow the procedures outlined above to calculate the main diagnostic quantities related to disk angular momentum transport and mass loss .
we first discuss the results from our fiducial set of physical parameters with @xmath157 , @xmath160 g @xmath152 , @xmath164 and @xmath165 , shown in figure [ fig : fiducial ] .
we focus our discussion on the solid lines in the figure , and in section [ sssec : shield ] we discuss the effect of wind shielding by comparing with the dashed lines in the figure .
the middle left panel shows the location of the fuv ionization front @xmath3 and wind base @xmath4 .
we can see that up to @xmath166 au , the two locations coincide with each other , meaning that the disk is largely laminar up to this radius , with wind being the dominant mechanism for angular momentum transport . the value of @xmath167 increases with increasing @xmath2 , meaning that the fuv ionization front traces a flared geometry . on the other hand , because the disk itself is flared , we see from the inset that when normalized to disk scale height , the value of @xmath64 decreases with increasing @xmath2 .
this means that the fuv radiation effectively penetrates deeper ( geometrically ) towards the outer disk .
the value of @xmath64 is @xmath168 at 1 au , and decreases to @xmath169 at @xmath170 au .
these numbers are similar to local disk simulations where the fuv penetration depth is measured in the vertical domain @xcite . deeper penetration of fuvs towards the outer disk allows the mri to operate at @xmath166 au and beyond in our fiducial model .
correspondingly , the wind base location @xmath4 no longer coincides with @xmath3 , and is located higher in the atmosphere . with @xmath64 decreasing with @xmath2 , the extent of the mri zone increases towards larger @xmath2 , and hence the value of @xmath41 increases towards the outer disk and beyond . nevertheless , because only a very small fraction of mass resides in the mri active zone , the averaged value @xmath41 remains relatively small ( @xmath171 ) .
we find that in the inner disk , the wind alfvn radius ( or lever arm ) decreases with increasing @xmath2 .
this is a consequence of @xmath172 getting smaller , and hence when assuming constant midplane @xmath115 , the wind base is less strongly magnetized ( larger plasma @xmath102 , or smaller @xmath173 , at the wind base ) . as a result , the wind becomes less efficient in transporting angular momentum towards larger @xmath2 , and the disk has to lose more mass to maintain its accretion rate ( i.e. , higher mass loading ) .
when the mri zone is present , because we have have set fixed @xmath118 at the wind base location , the alfvn radius and mass loading parameter become approximately constants ( as found in b16 ) .
the middle right panel of figure [ fig : fiducial ] shows the radial profiles of wind - driven accretion rate , viscous accretion rate , and wind mass loss rate per logarithmic radii .
we see that clearly , almost the entire disk relies on disk wind to transport angular momentum .
the only exception is near the disk outer edge , where disk surface density exponentially falls off and viscous transport picks up . because of the dominant role played by the wind , viscous disk spread is largely suppressed , except for regions well beyond the characteristic disk size of @xmath146 au .
although we have made an artificial assumption that midplane @xmath115 is constant throughout the disk , the overall wind - driven accretion rate is approximately constant over an extended range of disk radii , indicating that the system shall approximately maintain a steady state .
this is related to the our initial disk surface density that we have chosen , where the midplane pressure @xmath174 .
assuming that the wind - driven accretion rate depends quadratically on the field strength @xcite , a constant accretion rate would require @xmath175 , whose power law index is very close to @xmath176 .
the mass loss profile is best quantified by the mass loss rate per logarithmic radii @xmath177 , which is related to wind - driven accretion rate by ( see b16 ) @xmath178 as the lever arm @xmath179 ^ 2 $ ] decreases towards large @xmath2 , we see that wind mass loss rate rapidly increases with @xmath2 .
the mass loss rate per logarithmic radii is nearly two orders of magnitude below the accretion rate at the innermost radius , while it can become a significant fraction of the accretion rate at tens of au .
thus , we expect that most of the mass loss in ppds occurs through the outer disk via very slow winds , in contrast with observations which typically can only trace winds launched from the inner disk ( see section [ ssec : imp ] for further discussion ) .
we also note that magnetic field geometry from typical disk wind simulations ( e.g. , @xcite ) has larger inclination angle with respect to the disk at small disk radius ( @xmath180 ) , while @xmath74 becomes smaller towards the outer disk radii .
if we were to relax our assumption of constant field inclination @xmath75 , this would make @xmath22 larger in the inner disk and smaller in the outer disk ( though not by much in a warm wind , as discussed in b16 ) , and hence strengthen our conclusion that mass loss is most significant in the outer disk . both wind - driven accretion rate and wind mass loss rate drops when an mri zone is present .
this is can be qualitatively understood from the middle panel of figure 12 in b16 . with the wind launched from higher location hence lower density , both accretion and mass loss rates decreases , and the latter decreases faster .
it is this reduction of wind transport that leaves rooms for viscous effect to take over in the outermost region of the disk .
readers may note a bump at small @xmath2 in wind - driven accretion and mass loss rates in figure [ fig : fiducial ] .
this is due to the inability to properly compute fuv penetration near the inner boundaries , but it is only a minor effect and should not affect the overall calculations .
our calculations have incorporated the fact that a disk wind enhances gas density in the disk atmosphere compared with the hydrostatic case , using the approximate fitting formula ( [ eq : rhowind1 ] ) .
the enhanced density adds to the column density that fuv radiation has to penetrate , thus providing a certain level of shielding .
this shielding effect has been discussed in the context of survival of molecules through various stages of disk evolution @xcite .
@xcite further pointed out that strong disk outflow launched from the innermost region of ppds may suppress the wind launching process at larger radii by shielding fuv / x - ray photons . to assess the significance of this shielding effect
, we also perform a calculation that simply uses the hydrostatic density profile ( [ eq : rhomri ] ) at @xmath181 , and the results are shown in dashed lines in figure [ fig : fiducial ] .
we see that indeed , using hydrostatic density profile makes fuv radiation penetrate deeper , giving smaller @xmath3 at all disk radii .
although the absolute difference in @xmath3 values is small , it can be more clearly seen in the inset of the middle left panel . because density drops very rapidly with height in a gaussian density profile , a small difference in @xmath3 translates to much larger difference in the density @xmath182 there . without shielding ,
we see that the wind mass loss rate becomes larger by a factor of nearly two .
this is because launching the wind from higher density means that the wind is effectively less magnetized , and hence become more heavily loaded ( see again from the middle panel of figure 12 in b16 ) .
thus , the presence of the wind itself helps prevent more severe wind mass loss from the disk ( at a modest level ) .
we also note that the wind launched from the thermally ionized , fully turbulent innermost disk , can also contribute to this shielding effect , which is not included in our calculations ( see , e.g. , @xcite ) .
in addition , we also see that without shielding , the mri active zone in the outer disk becomes both radially broader and vertically thicker . at each radius ,
a higher fraction of disk mass is located in the mri active zone , resulting in a larger @xmath41 .
we discuss in this subsection the role played by individual physical parameters , including total magnetic flux ( parameterized by midplane @xmath115 ) , fuv penetration depth ( @xmath59 ) , and atmosphere temperature ( characterized by @xmath183 ) .
we vary one parameter at a time .
the results are shown in figures [ fig : param ] and [ fig : snapshots ] .
note that figure [ fig : snapshots ] is exactly analogous to figure 12 of b16 , but applied to a full disk model . in this figure ,
we quote accretion rate measured at 1 au ( as discussed earlier , accretion rate profile is largely flat given our magnetic flux prescription ) . for mass loss , we quote the enclosed mass loss rate ( [ eq : mloss ] ) within radius of @xmath184 au , @xmath185 au and infinity , respectively , to exemplify the mass loss profile .
the left panels of figure [ fig : param ] show the effect of enhanced net vertical magnetic flux , where we choose @xmath186 at midplane . because density structure in the vicinity of the wind base does not change with magnetization ( see section [ sssec : den ] ) , the location of @xmath3 remain largely unchanged . due to stronger magnetization ,
the outer disk is less prone to the mri and the mri active zone shrinks .
while the @xmath0 value is larger in the disk interior at all radii to reflect potentially enhanced viscous transport ( section [ sssec : vis ] ) , and yields higher viscous - driven accretion rate , wind plays a more dominant role in global transport : even in the disk outer edge , wind - driven accretion rate overwhelms viscous decretion rate , and hence viscous spread is suppressed .
stronger vertical fields leads to much enhanced wind - driven accretion rate , but only a modest increase in wind mass loss rate .
this effect is better reflected in the left panel of figure [ fig : snapshots ] , and has also been extensively discussed in b16 . for our fiducial parameters ( midplane @xmath157 ) ,
the mass loss rate integrated over the entire disk is about the same as the mass accretion rate
. the ratio drops to @xmath187 and @xmath188 for midplane @xmath186 and @xmath189 respectively .
thus , level of disk magnetization largely controls the fractional mass loss rate over accretion rate .
in addition , we see that mass loss from within @xmath184 au represents only a very small fraction of total mass loss rate , about @xmath190 .
this fraction increases to about @xmath191 for mass loss enclosed with @xmath192 au .
thus , we reinforce the conclusion that most of the mass loss is achieved at the outer disk , which is probably not easily observable . the middle panels of figure [ fig : param ] shows the effect of enhanced fuv penetration depth @xmath193 g @xmath152 .
deeper penetration clearly lowers the location of @xmath3 . in the inner disk ( before the mri active zone appears ) , the wind base density and pressure becomes higher , making the wind less strongly magnetized . as discussed in b16 , this leads to the development of stronger toroidal field , which drives both higher wind - driven accretion rate , and higher wind mass loss rate . with a reduction of the alfvn radius , the mass loss rate increases more rapidly .
this trend is more clearly seen in the middle panel of figure [ fig : snapshots ] , and is analogous to the middle panel of figure 12 in b16 .
this is a manifestation of what we termed magneto - photoevaporation " in local wind study of b16 to a global disk model . in the mean time
, deeper fuv penetration enlarges the mri active zone both radially and vertically , leading to enhanced viscous angular momentum transport . from figure
[ fig : param ] , we see that viscous transport starts to dominate over the wind at a smaller radius than in our fiducial case . nevertheless , this radius is still larger than our model disk size of @xmath146 au , and hence wind transport still dominates the bulk of disk evolution .
we also note that our calculation have ignored external fuv irradiation , which likely dominates over fuv from the star at a few tens to @xmath170 au scale @xcite , depending on the intensity of the external radiation field and stellar fuv luminosity . such external irradiation ,
if sufficiently strong , likely allows fuv to penetrate deeper into the outer disk because they are not shielded by the inner disk . correspondingly , we would expect the wind to become less effective in the outer disk , where viscous transport likely play a more significant role . ) using our fiducial model parameters .
lines with different transparencies represent different times , as indicated by the legend .
black solid lines correspond to the case where total magnetic flux @xmath106 is set to be proportional to disk mass @xmath194 , while red dashed lines represent the case where @xmath106 is conserved . for comparison
, we also show in blue thin dash - dotted lines results from a pure viscous evolution calculations assuming constant @xmath195 at t=1.0 , 2.0 and 4.0 myrs.,width=340 ] disk atmosphere temperature mainly affects wind properties in two ways .
first , it directly affects wind properties via thermodynamics , as studied in detail in b16 .
second , it modifies the surface density structure . in our two - temperature disk model
, a hotter disk surface first leads to a smaller density right above the fuv front due to pressure equilibrium at @xmath45 .
on the other hand , @xmath80 decreases with @xmath196 much more slowly at @xmath181 .
these two effects turn out to approximately cancel and only weakly affect the geometric depth of fuv penetration , as seen from the right panel of figure [ fig : param ] . assuming
the atmosphere temperature does not affect the wind base location , b16 found that the overall wind - driven accretion rate and wind mass loss rate are relatively insensitive to variations in wind temperature . incorporating the changes in surface density structure ,
our global calculations show that the same conclusion holds .
further in figure [ fig : snapshots ] , we see that both mass accretion rate and total mass loss rate are very insensitive to @xmath183 .
in this section , we explore the long - term disk evolution based on prescriptions described in section [ ssec : model ] .
we here mainly focus on the global mass budget , in terms of evolution timescale , fractional mass loss through accretion and wind , etc .
, to minimize uncertainties associated with the magnetic flux distribution and evolution .
in figures [ fig : sigevolve ] and [ fig : rateevolve ] , we show the time evolution of disk surface density , and the associated disk mass ( @xmath194 ) , accretion rate and mass loss rate evolution , where accretion rates are measured at the inner disk boundary .
two scenarios of magnetic flux evolution are considered , namely , either total magnetic flux @xmath159 , or @xmath106 is constant .
we see that at early stages , disk evolution proceeds very similarly in these two scenarios .
after about @xmath184 myrs , the disk has lost about @xmath197 of its mass through accretion and wind , and the subsequent evolutionary paths in the two scenarios diverge .
when assuming @xmath159 , the accretion rate gradually decreases with time as magnetic flux is lost , and the evolution slows down . even after @xmath198 years of evolution , the disk still possesses about @xmath199 of its initial mass .
we also see that midplane @xmath115 slowly increases with time , and towards later evolution , wind mass loss rate exceeds accretion rate . when assuming @xmath106 is constant , on the other hand , we see that accretion rate and mass loss rate maintain approximately a constant level for @xmath91 myrs .
this is because magnetic field strength roughly stays constant in this scenario , which directly regulates accretion rates .
constant accretion / mass loss rates lead to rapid depletion of disk materials , and we see that total disk mass plunges down in a runaway manner shortly after @xmath200 myrs . as the majority of disk mass is accreted or lost , fuv radiation can penetrate substantially deeper , enhancing the mass loss process relative to accretion . in the mean time , the mri active zone grows and allows viscous spread to dominate in the outermost disk .
the reduction of accretion rate near the end is a result of the redistribution of magnetic flux towards the outer disk ( due to viscous spreading ) .
the two scenarios discussed here can be considered as two extreme limits on magnetic flux evolution .
rapid loss of magnetic flux leads to slow evolution and long disk lifetime , while conservation of magnetic flux leads to a two - timescale behavior : disk depletion occurs on timescales much shorter than disk lifetime .
the latter behavior was also discussed by @xcite when assuming constant total magnetic flux .
the reality may lie in between the two extreme scenarios , which we use in the next subsection as a way to constrain disk lifetime . for comparison
, we also show in dash - dotted lines of figure [ fig : sigevolve ] results from a pure viscous disk evolution model assuming constant @xmath195 .
besides the sequential drop in surface density , disk evolution is characterized by the expansion of the outer disk ( viscous spreading ) . over the course of a few myrs , the disk size ( defined by the radius above a certain threshold surface density )
have expanded by more than a factor of 2 .
this is much more significant than that in our fiducial disk evolution models , even viscous spreading dominates in the outermost region of the disk .
thus , we conclude that wind - dominated ppd evolution likely undergoes very little expansion of the outer disk as compared with viscous evolution models . to better quantify the timescale of disk evolution and the significance of mass loss from disk winds , we have carried out a series of disk evolution calculations scanning the parameter space . because we have found that the wind transport properties are the least sensitive to wind temperature @xmath68 , we fix @xmath201 in all the calculations .
we consider midplane plasma @xmath202 , @xmath154 , @xmath155 and @xmath156 , and fuv penetration depth @xmath160 and @xmath203 g @xmath152 .
we do not evolve the disk all the way to the end , but terminate the evolution when the disk has lost half of its mass , and call this time half disk lifetime @xmath204 .
we do so mainly because long - term evolution calculations likely bare large uncertainties due to our ignorance on magnetic flux evolution . for each set of parameters
, we perform two runs , evolving the disk either assuming magnetic flux conservation @xmath158 const , or @xmath159 .
as discussed earlier , we may consider these two cases as two extreme scenarios , which set the lower and upper limits of @xmath204 .. based on our results , if magnetic flux is approximately conserved during disk evolution , we approximately have @xmath205 . if the disk loses magnetic flux , then @xmath206 .
] similarly , we evaluate @xmath207 , the fractional mass loss from the wind compared to total mass loss due to both wind and accretion , over evolution time up to @xmath204 . in figure
[ fig : lifetime ] , we plot @xmath204 and @xmath207 from all runs in this parameter study , organized as a function of initial midplane @xmath115 .
it best summaries the main results discussed in this paper .
disk lifetime is determined by a combination of mass accretion and mass loss processes .
it mostly depends on the amount of magnetic flux threading the disks . for our fiducial disk model , midplane plasma @xmath208 yields @xmath209 that is in best agreement with observational constraints of disk lifetime ( e.g. , @xcite ) on a few myr time scale .
it also supports the interpretation of fossil magnetic field measurement of the solar nebular @xcite .
note that increasing or reducing the field strength by a factor of only @xmath210 ( a factor @xmath192 change in @xmath115 ) would yield disk lifetime that is either too short or too long .
all calculations with @xmath157 yield a fractional wind mass loss @xmath207 around one half ( 0.3 - 0.7 ) .
wind mass loss becomes progressively less important towards stronger disk magnetization , but such scenario is very unlikely as constrained by disk lifetime .
therefore , we expect that wind mass loss plays a rather significant role in global disk evolution .
enhanced fuv luminosity ( hence its penetration ) leads to both enhanced accretion rate , and to a higher level , outflow rates ( and hence larger @xmath207 ) . in our fiducial model
, @xmath207 increases from from 0.45 to 0.65 as @xmath59 increases from @xmath211 to @xmath203 g @xmath152 , which manifests what we termed magneto - photoevaporation in b16 .
the contrast between mass accretion and outflow is more prominent as the disk becomes more strongly magnetized . on the other hand , because of the relatively gentle dependence of @xmath7 and @xmath212 on @xmath59 ( see figure [ fig : snapshots ] ) , reduction of @xmath204 is only modest .
a factor of @xmath185 difference in @xmath59 typically leads to a factor of at most @xmath213 difference in @xmath204 .
in this subsection , we consider disks that are less flared , which is expected to affect the penetration of fuv / x - rays in heating / ionizing the disk surface layer . although the temperature structure of ppds is not well constrained observationally , this study is in part motivated by the sed - based classification of herbig ae / be disks that show two distinct groups with either flared or flat geometry @xcite , it has recently been found that flared herbig ae / be disks all have large gaps @xcite , and some flat disks may have small gaps @xcite , suggesting a evolutionary path starting from flat full disks towards flared transition disks .
note that this classification mainly reflects disk geometry at @xmath214au scale , where flat disks are generally attributed to self - shadowing by puffed inner rims @xcite .
even many disks are considered
flat " , they can become flared towards larger radii ( e.g. , @xcite for the hd 163296 disk ) . despite the complications and uncertainties in our understandings of disk geometry and temperature structure , we aim to assess whether they can significantly affect disk wind / transport properties in a qualitative manner by simply varying the parameter @xmath49 .
note that @xmath215 would lead to a constant @xmath216 , which is unlikely to be realized on a global scale when balancing passive stellar heating with thermal radiation . in figure
[ fig : flare ] , we compare the results from our fiducial model with normal level of flaring ( fiducial , @xmath165 ) , a model with less flaring ( @xmath217 ) , and an almost flat disk ( @xmath218 ) .
note that increasing @xmath49 makes the disk cooler towards the outer disk , with smaller disk scale height @xmath52 .
correspondingly , the geometric location of the fuv front @xmath3 becomes lower . on the other hand , we see from the left panel of the figure that when normalized to disk scale height , @xmath64 becomes progressively higher as the disk becomes less flared .
this is simply a geometric effect that at the same @xmath219 , sight lines to the central star would encounter larger column density of gas in a less flared disk . in addition , we see that in less flared disks , the size of the mri active zone shrinks .
this is also natural consequence of larger @xmath64 . with @xmath217 ,
the inner edge of the mri active zone is already at @xmath220 in our disk model , while for a nearly flat disk with @xmath218 , the mri active zone diminishes : the entire disk becomes largely laminar . with larger @xmath64 in less flared disks , the wind base density and pressure is smaller , making the wind base region effectively more strongly magnetized .
as shown in the right panel of figure [ fig : flare ] , the alfvn radius becomes larger in less flared disks .
this means that wind mass loss rate becomes a smaller fraction of accretion rate ( see equation ( [ eq : lever ] ) ) .
this is also reflected in figure [ fig : lifetime ] , where we also show results of @xmath204 and fraction mass loss with @xmath217 .
overall , the situation for less flared disk is similar to the case with smaller @xmath59 : less flared disks lead to a reduction of both wind - driven accretion rate and mass loss rate ( and hence longer @xmath204 ) , but the reduction of mass loss rate is more significant .
the studies in the preceding sections have implicitly assumed that x - rays and fuv have similar penetration depth , as described in section [ ssec : model ] . in reality ,
fuv penetration depth is very uncertain depending on the abundance of very small grains , while absorption of x - ray photons is insensitive to grain abundance , and depends mainly on photon energy @xcite . for typical photon energy of @xmath184 kev ,
the x - ray penetration depth is on the order @xmath221 @xmath152 ( e.g. , @xcite ) , and it becomes larger for higher photon energies .
this is at least comparable to the fiducial value of @xmath160 g @xmath152 that we have adopted .
however , considering the full x - ray spectrum with the presence of harder x - rays , the overall penetration depth is likely higher . in this subsection
, we relax our previous assumption and allow x - rays to penetrate deeper than the fuv .
we define x - ray penetration depth @xmath222 , which determines the location of atmosphere @xmath3 .
gas temperature is set to @xmath43 for @xmath79 and @xmath44 for @xmath181 , as before .
may no longer be a good approximation .
however , we mainly aim at a proof - of - concept study on the role of deeper x - ray heating , and a two - temperature treatment should suffice for this purpose .
] let @xmath223 be the height of the fuv ionization front set by @xmath59 , and @xmath224 .
the wind base is assumed to be located at @xmath225 if gas is sufficiently magnetized to suppress the mri , otherwise , @xmath4 is set to larger values according to the prescription in section [ sssec : base ] . in figure
[ fig : xray ] , we compare the results from our fiducial model with @xmath226 g @xmath152 ( solid lines ) with a calculation assuming @xmath227 g @xmath152 ( dashed lines ) .
the left panel shows the value of alfvn radius normalized to wind launching radius .
we see that despite that x - rays can heat into deeper regions in the disk , leading to substantial differences in disk vertical density and temperature structures , the wind properties resulting from the two cases are remarkably similar .
this is most easily understood from the example in the right panel , where we show the vertical pressure profiles at 1 au in the two cases .
with @xmath44 at disk surface layer , the density profile , as well as pressure profile , are in the form of @xmath228 .
if we assume the column densities ( from a position at disk surface ) radially towards the star and vertically towards large @xmath196 are proportional to each other ( which approximately holds for flared disks where column density is dominated by local contributions ) , then for a given @xmath59 , the density and pressure at the fuv ionization front @xmath223 ( @xmath229 in this case ) are largely unchanged regardless of whether x - rays penetrate deeper or not . in this example , we find the pressure ( and density ) difference between the two cases at @xmath230 is only @xmath231 , which explains the very small difference in @xmath26 .
therefore , we conclude that density and pressure at the wind base ( and hence the main wind properties ) is largely insensitive to additional heating by x - rays below the fuv ionization front . we have also verified this result using higher atmosphere temperature @xmath232 , and with larger @xmath233 up to @xmath234 g @xmath152 .
photoevaporation has long been considered as the primary mechanism for the dispersal of ppds . as a pure thermal wind
, it does not exert a torque to the disk , and hence would evaporate the disk regardless of internal processes of disk angular momentum transport .
as mhd wind has been realized to play the dominant role in disk angular momentum transport , we expect real ppd winds to be magnetized in nature , and they are externally heated . therefore , wind mass loss and angular momentum transport are intrinsically coupled .
they are both affected by thermal effects , and they jointly determine global disk evolution . incorporating the magneto - thermal disk wind model of b16 ,
our model has included , in a simple approximate manner , all major ingredients of the aforementioned new framework of global disk evolution . compared with photoevaporation calculations ,
wind mass loss in our calculations is directly coupled to angular momentum transport , instead of being treated separately .
this leads to the correlation between disk lifetime and fractional wind mass loss , as a natural consequence of mhd disk wind .
photoevaporation calculations are highly sensitive to the treatment of thermodynamics , which leads to considerable discrepancies between different models such as those driven by extreme uv ( e.g. , @xcite ) , x - rays ( e.g. , @xcite ) , and fuv ( e.g. , @xcite ) .
while our calculations have treated thermodynamics in a highly simplified manner , our main results suggest that wind mass loss is only modestly sensitive to thermodynamic effects , whereas the dominant role is played by amount of magnetic flux .
as suggested in b16 , a unified description of disk wind mass loss would be better termed magneto - photoevaporation " .
photoevaporation has been commonly invoked as an explanation of ( a fraction of ) transition disks , a small fraction of ppds characterized by the presence of large inner gaps / holes , by inside - out clearing ( see @xcite for a review ) . in the mhd framework , whether the disk can be cleared inside - out depends on the details of magnetic flux evolution , which remains poorly understood .
for instance , if the inner disk is capable of maintaining its total magnetic flux ( e.g. , possibly due to the hall effect ) , while the outer disk loses flux ( i.e. , as a result of ambipolar diffusion ) , then inside - out clearing would be a natural consequence .
one important implication of wind mass loss is that mass is primarily removed from disk surface , while in low turbulent environment , most dust / solids settle and reside around disk midplane .
the removal of largely dust - free gas , and in combination with radial drift ( e.g. , @xcite ) , can directly enhance the dust - to - gas mass ratio , creating favorable conditions for planetesimal formation @xcite , and particularly allowing for smaller , more strongly coupled dust to participate in planetesimal formation @xcite .
dust evolution has been incorporated in the recent study by @xcite in the context of photoevaporation on top of viscous disk evolution , and they found significant enhancement of dust - to - gas ratio as a robust outcome of surface mass loss .
we thus expect the same conclusion to hold in the mhd wind , or magneto - photoevaporation framework . the fact that wind mass loss is most significant towards the outer disk implies that the enhancement of dust - to - gas ratio may proceed from the outer disk inward .
signatures of wind from ppds have been routinely observed in the form of blue - shifted emission lines such as from co , oi and neii lines ( e.g. , @xcite ) , but constraining the wind mass loss rate from line data is extremely difficult .
is found to be of the order @xmath203 with large uncertainties @xcite . ]
@xcite recently reported @xmath235 in the range of @xmath236 with large uncertainties . with typical blueshift of the order a few km s@xmath237 and given the excitation conditions , the wind most likely originates from the inner few au .
our results suggest that the inferred mass loss rate only represents a small fraction ( @xmath238 ) of total disk wind mass loss . within the uncertainties ,
the measurements are consistent with our expectations of @xmath239 .
we have discussed in section [ sssec : shield ] that the column density in the wind flow can provide shielding for the uv radiation from the protostar .
wind shielding has also been discussed in the context of survival of molecules @xcite . in reality ,
fuv penetration is largely determined by the abundance of very small grains in the disk surface and wind column , which is very poorly known .
we speculate that if very small grains are sufficiently abundant at disk surface , they can be efficiently lifted by the wind ( e.g. , @xcite ) , which can substantially enhance the fuv opacity .
the non - detection of turbulent motion towards the outer region of herbig ae disk hd 163296 @xcite also indirectly suggests that shielding of fuv radiation may reduce or even quench the mri zone in the outer disk .
our fiducial choice of a relatively small @xmath59 reflects such considerations , but the potentially important effect of wind shielding calls for further study .
our global model inherits from the wind model of b16 , thus share the same uncertainties , especially from the fact that we prescribe the wind geometry instead of solving cross - field force balance .
however , as discussed there , more uncertainties arise from our ignorance about magnetic flux distribution , which eventually determines the wind geometry .
thus , b16 chose to parameterize the disk geometry and focus on wind physics along prescribed field lines .
our global wind model follows the same logic , and mainly explores the consequence of the b16 model at a global scale .
our wind model may suffer from two systematic uncertainties .
the first arises from a potentially systematic variation of wind geometry with radius .
for instance , global wind simulations ( e.g. , @xcite ) typically find that towards larger radii , the wind inclination angle @xmath74 increases , and wind field lines get more divergent .
we have adopted fixed wind geometry parameters @xmath75 and @xmath76 in this work . based on results from the b16 model , with field lines getting
more inclined and more divergent , the alfven radius ( lever arm ) would decrease , leading to more pronounced fractional mass loss towards the outer disk .
the second uncertainty arises from the extension of the b16 wind model to wind launched from the mri active region , whereas the model is intended for a laminar wind .
wind launching in the mri case is a robust phenomenon as observed in local simulations @xcite , and the field configuration is fully dominated by the toroidal component indicating wind launching by magnetic pressure gradient , similar to the laminar wind in ppds discussed in b16 .
we thus expect our treatment at least qualitatively reflects the relation between mass loss and angular momentum transport in this regime , while global simulations are certainly needed to better calibrate wind kinematics .
our calculations parameterize the role of thermodynamics simply in terms of fuv ( and x - ray ) penetration , constant wind temperature and level of disk flaring . in reality , radiative transfer and photo - chemistry are essential to better quantify thermodynamical effects . due to the important role played by grain opacity ,
these are coupled with the size distribution and spatial distribution of dust grains , which in turn are coupled to disk dynamics .
the interplay among dynamics , dust evolution , chemistry , and radiation in the system may result in much richer phenomenology than considered here , and are worth more in - depth investigations .
the sensitive dependence of wind and transport properties on magnetic field strength calls for better understandings of magnetic flux transport , as already discussed in section [ sssec : bz ] .
moreover , the starting point of our evolution calculations is motivated from a typical disk at class ii phase .
disk evolution starts from much earlier phases , and magnetic flux is well known to control disk dynamics from the disk formation stage ( see @xcite for a review ) .
in particular , the hall effect could induce a bimodality on initial disk size depending on the polarity of the background magnetic field with respect to disk rotation axis @xcite , and such bimodality is likely to be carried into all later phases of disk evolution @xcite .
how the hall effect affect magnetic flux evolution is unknown , but hints for strong polarity dependence exist from local studies @xcite .
our figure [ fig : lifetime ] shows that to reproduce observed ppd lifetime , the amount of magnetic flux is constrained to a relatively narrow range .
some feedback and self - regulation mechanism might be involved in magnetic flux evolution .
future global simulations of wind - driven disk evolution with resolved disk microphysics are essential to address these problems .
in this work , we have constructed a framework to study global evolution of ppds that incorporates wind - driven accretion , wind mass loss , and viscous transport . the magneto - thermal wind model the b16
is adopted as the primary ingredient of the framework .
the model is motivated by recent local simulations of ppd gas dynamics that have properly incorporated disk microphysics , which suggest that the mri is largely suppressed in the disk interior and magnetized wind is launched in the externally heated and ionized surface layer from the fuv ionization front @xcite .
we further consider the fact that the well ionized surface layer can be subject to the mri in the outer region of ppds @xcite , depending on the strength of vertical magnetic field fields and fuv penetration depth , which contributes to viscous angular momentum transport .
ideally , disk evolution should be coupled with a procedure for magnetic flux evolution .
the latter is treated phenomenologically in this work by assuming constant midplane plasma @xmath115 , but can be improved upon better understanding of magnetic flux transport in ppds .
* disk evolution is largely dominated by mhd wind - driven accretion and mass loss . contribution from the mri
can be important in the outermost disk but viscous spread is suppressed . *
the disk evolution timescale sensitively depends on the amount of magnetic flux threading ppds .
disk dispersal is rapid if the disk is able to retain most of its magnetic flux during evolution , otherwise , disk dispersal is gradual .
* given typical disk lifetime of a few myrs , the disk loses comparable amount of mass via disk wind and accretion .
most of the wind mass loss proceeds through the outer disk ( @xmath240 au ) .
fractional mass loss via disk wind increases with decreasing disk magnetization ( increasing disk lifetime ) . *
the depth of fuv penetration and level of disk flaring are the main thermodynamic factors affecting disk evolution .
smaller fuv penetration depth and less flared geometry slightly reduce wind - driven accretion rate , and more strongly reduces fractional wind mass loss .
this work represents an initial effort towards modeling global evolution of ppds that incorporates realistic disk physics . at the moment
, we have mainly focused on the evolution of the overall mass budget , and significant wind mass loss likely substantial enhances the disk dust - to - gas mass ratio and directly promotes planetesimal formation .
better knowledge on magnetic flux evolution is needed to make reliable predictions on the details of disk surface density evolution , allowing for more realistic studies of planet formation .
i thank kees dullemond and antonella natta for encouraging me to work on this project , and sean andrews for useful discussions .
i also thank jeremy goodman , antonella natta , kees dullemond , and an anonymous referee for carefully reading of the manuscript and providing useful feedback that improves this paper .
this work is supported by institute for theory and computation at harvard university .
, r. r. , weiss , b. p. , lima , e. a. , harrison , r. j. , bai , x .-
n . , desch , s. j. , ebel , d. s. , suavet , c. , wang , h. , glenn , d. , le sage , d. , kasama , t. , walsworth , r. l. , & kuan , a. t. 2014 , science , 346 , 1089 | a global evolution picture of protoplanetary disks ( ppds ) is key to understanding almost every aspect of planet formation , where standard alpha - disk models have been constantly employed for its simplicity . in the mean time
, disk mass loss has been conventionally attributed to photoevaporation which controls disk dispersal .
however , a paradigm shift towards accretion driven by magnetized disk winds has been realized in the recent years , thanks to studies of non - ideal magneto - hydrodynamic effects in ppds .
i present a framework of global ppd evolution aiming to incorporate these advances , highlighting the role of wind - driven accretion and wind mass loss .
disk evolution is found to be largely dominated by wind - driven processes , and viscous spreading is suppressed .
the timescale of disk evolution is controlled primarily by the amount of external magnetic flux threading the disks , and how rapidly the disk loses the flux .
rapid disk dispersal can be achieved if the disk is able to hold most of its magnetic flux during the evolution .
in addition , because wind launching requires sufficient level of ionization at disk surface ( mainly via external far - uv radiation ) , wind kinematics is also affected by far - uv penetration depth and disk geometry . for typical disk lifetime of a few myrs ,
the disk loses approximately the same amount of mass through the wind as through accretion onto the protostar , and most of the wind mass loss proceeds from the outer disk via a slow wind .
fractional wind mass loss increases with increasing disk lifetime .
significant wind mass loss likely substantially enhances the dust to gas mass ratio , and promotes planet formation . |
imagine a stable hamiltonian system that is subjected to an external periodic perturbation as well as a frictional force that continuously takes energy out of the system .
if the driving force vanishes on the average , then in general the average behavior of the system is characterized by damped motion . under favorable conditions ,
however , the system could get captured into a prolonged resonance if the driving frequency is commensurate with an internal frequency of the system . in this paper , we study the dynamics of this phenomenon in the context of a simple model of certain relativistic binary systems that are of current interest .
relativistic binary pulsars are generally expected to occur in some astrophysical environment and the gravitational effect of the surroundings on the binary orbit would be tidal in the first order of approximation . the evolution of the binary orbit under the combined influence of gravitational radiation reaction and the external tidal perturbation can lead to resonance under appropriate conditions [ 16 ] .
once we average over the external periodic tidal perturbation , the binary pulsar is expected to lose energy to gravitational radiation damping so that the relative orbit would spiral inward unless there is resonance capture .
this would come about if the external source deposits energy into the orbit over an extended period of time so as to compensate for the continuous loss of energy due to radiation reaction .
this situation , which persists until the system leaves the resonance , can occur if the keplerian period of the osculating ellipse becomes commensurate with the period of the external perturbation . in this way
, the binary pulsar can be captured into resonance with its gravitational environment .
then , the standard data reduction schemes are expected to reveal that during resonance capture the semimajor axis of the binary is not monotonically decreasing due to the emission of gravitational radiation but is instead oscillating about a fixed resonance value since the external perturbation continuously deposits energy into the orbit during resonance so as to replenish on average the energy lost to gravitational waves .
the system leaves the resonance when this delicate energy balance is significantly disrupted .
such resonances are well known in the solar system , where the damping effects have a different origin ( see @xcite and the references cited therein ) .
one expects that over time the same kind of phenomenon will be observed in a sufficiently large sample of relativistic binary pulsars @xcite .
it should be mentioned that there are transient chaotic phenomena that are generally associated with resonance . the transient chaos _
per se _ may be difficult to detect observationally due to noise and other complications that exist in the data reduction process @xcite .
on the other hand , there is numerical evidence for a peculiar chaotic effect , namely _ chaotic transition _ , that appears to be associated with transient chaos and involves a sudden transformation in the rate of orbital collapse @xcite .
that is , the system makes a transition from one relative orbit to another that collapses much more rapidly .
presumably such a chaotic transition if it indeed occurs in nature should be detectable by the timing observations of relativistic binary pulsars . though the unperturbed binary systems in our investigations always involve standard astronomical masses on keplerian orbits , it is possible to extend the analysis to geodesic orbits around black holes .
such general relativistic systems involve stable kepler - like orbits as well as unstable orbits .
we expect that our analysis of transient chaos could be extended to the stable orbits , while the occurrence of chaos involving the unstable orbits perturbed by incident gravitational radiation has been demonstrated by bombelli and calzetta @xcite
. these results could be significant for the study of chaotic accretion onto black holes and possible chaotic enhancement in the emission of gravitational waves associated with active black holes .
the purpose of the present work is to show that resonance capture can indeed occur on average for a simplified model of the physical situation that has been described here , since our previous work on sustained resonance has been mainly numerical [ 2,4,5 ] .
our mathematical model is briefly discussed in the next section .
the second - order partially averaged system is given in section [ sec : pave ] .
the geometry of resonance capture for the second - order partially averaged system is presented in section [ sec : gorc ] , and it is shown that resonance capture can indeed occur for this partially averaged model under certain conditions .
the proof of a resonance capture mechanism for the full model equations remains an important unsolved problem ; instead , we provide numerical evidence for sustained resonance in our model in section [ sec : discuss ] .
in a series of papers [ 16 ] , we have studied the long - term evolution of a keplerian binary system coupled to a gravitational environment .
in fact , the influence of the environment has been replaced for the sake of simplicity and definiteness by the tidal perturbation of an external gravitational wave .
more precisely , we have considered the _
simplest _ gravitational model for a binary system under radiative perturbations . in the lowest ( quadrupole ) approximation , the external wave exchanges energy and angular momentum with the binary system but not linear momentum . for the purposes of the present investigation , we limit our considerations to a normally incident plane monochromatic external wave that is circularly polarized .
the amplitude of this external perturbation is @xmath0 , @xmath1 ; it indicates the deviation of the spacetime metric from the minkowski metric as a consequence of the existence of the incident radiation field . in practice
, @xmath0 could be extremely small ; in fact , direct laboratory detection of gravitational waves from astrophysical sources with @xmath2 is the goal of current experimental efforts .
the relative motion of the binary is planar due to the transversality of the external wave and is given by @xmath3 where @xmath4 determines the relative orbit , the indices @xmath5 , @xmath6 , and @xmath7 . here
the binary system consists of two point masses @xmath8 and @xmath9 with the center of mass of the system at rest at the origin of spatial coordinates , i.e. @xmath10 and @xmath11 .
the radiation reaction term is given to lowest order by @xmath12,\ ] ] where @xmath13 and @xmath14 describe the relative motion in terms of polar coordinates in the orbital plane and an overdot indicates differentiation with respect to @xmath15 .
the influence of the external radiation on the relative motion is represented in equation ( 1 ) by a symmetric and traceless tidal matrix @xmath16 evaluated at the position of the center of mass , @xmath17 , \end{aligned}\ ] ] where @xmath18 is the frequency of the external wave and @xmath19 , which is of order unity , is its amplitude .
the upper ( lower ) sign indicates right ( left ) circularly polarized radiation in the transverse - traceless gauge .
a general comment is in order here regarding the physical realizability of such a perturbation .
we can , for instance , imagine another binary system with a relative orbit that is _ circular _ with keplerian frequency @xmath20 .
then , in the rest frame of this binary s center of mass and neglecting any dissipation the system emits gravitational radiation of frequency @xmath18 and the waves that propagate perpendicularly to the orbital plane of the circular binary system are circularly polarized ; far from the circular binary , these waves are nearly planar and of the form under consideration here . in the absence of damping , we have shown that if @xmath0 is sufficiently small , the planar relative motion is bounded for all time as a consequence of the kolmogorov - arnold - moser ( kam ) theorem @xcite .
the physical reason for this confinement is that the external wave does not monotonically deposit energy into the binary orbit ; in fact , the general situation is that energy flows back and forth between the wave and the binary so that on average no net transfer of energy takes place .
when the orbital damping is taken into account , the general tendency of the binary system is to collapse ; however , under certain circumstances the system could get captured into resonance .
the necessary condition for resonance is the commensurability of the keplerian frequency @xmath21 and the wave frequency @xmath18 ; that is , relatively prime integers @xmath22 and @xmath23 must exist for which @xmath24 . sustained resonance would actually come about if a delicate balance could be established between the radiative perturbations .
it is useful to express the equations of motion in dimensionless form as in our previous work [ 35 ] . to this end ,
we express all spatial intervals in units of @xmath25 and all temporal intervals in units of @xmath26 , where the scale parameters @xmath27 and @xmath28 are connected via @xmath29 .
specifically , we assume that at some `` initial '' time the lengthscale @xmath27 is the semimajor axis of the binary orbit and @xmath30 is its period . in terms of polar
coordinates @xmath31 , the equations of motion then take the form @xmath32 \dot{\cal p}_{\vartheta } = 2 \delta \frac{{\cal p}_{\vartheta}}{\rho^{3 } } \left ( 9 { \cal p}_{\rho}^{2 } - 6 \ ; \frac{{\cal p}_{\vartheta}^{2}}{\rho^{2 } } + \frac{2}{\rho } \right ) + \epsilon \alpha \omega^{2 } \rho^{2 } \sin { ( 2 \vartheta \mp \omega t)}. \end{aligned}\ ] ] here @xmath33 is the dimensionless parameter that characterizes radiation damping and is given by @xmath34 for physical systems , @xmath35 @xcite .
we find that @xmath36 for the hulse
taylor relativistic binary pulsar @xmath37 ; this same value for @xmath33 is also approximately valid for the relativistic binary pulsar @xmath38 @xcite . to study the general characteristics of system , it is useful to express this system in terms of delaunay s elements .
these are the natural action - angle variables for the kepler problem . at each instant of time
@xmath15 , the state of relative motion @xmath39 , or equivalently @xmath40 , defines an osculating ellipse that is tangent to the perturbed motion at @xmath15 .
the delaunay variables @xmath41 are closely related to the orbital elements of the osculating ellipse .
let @xmath42 and @xmath43 be the semimajor axis and the eccentricity of the osculating ellipse , respectively ; then , @xmath44 , @xmath45 , and @xmath46 , where @xmath47 is the true anomaly of the osculating ellipse .
moreover , let @xmath48 be the corresponding eccentric anomaly ; then , the delaunay elements are defined by @xcite @xmath49 only positive square roots are considered throughout this work .
thus we limit our considerations to orbits with positive orbital angular momentum @xmath50 .
the delaunay equations of motion can be written as @xmath51 where @xmath52 , @xmath53 , for @xmath54 , are radiation damping terms @xcite , and @xmath55 \big\}. \end{aligned}\ ] ] here @xmath56 , \nonumber \\ b_{\nu}:=-\frac{8}{\nu^{2}e^{2}}(1-e^{2})^{1/2}\:[ej_{\nu}^{\prime}(\nu e)-\nu ( 1-e^{2})j_{\nu}(\nu e ) ] , \end{aligned}\ ] ] where @xmath57 is the bessel function of the first kind of order @xmath58 and @xmath59 . finally , we note that @xmath60 will be viewed as a fixed constant ; that is , we will only consider perturbations in the @xmath61 parameter space along lines with equations of the form @xmath62 .
let us note that in system there is only one `` fast '' angle , namely the mean anomaly @xmath63 , and its frequency is the kepler frequency @xmath64 of the osculating ellipse .
since @xmath65 is explicitly time dependent , resonance occurs when @xmath66 , in which case @xmath67 becomes fixed and equal to @xmath68 .
the dynamics of the system at the @xmath69 resonance is discussed in the next section .
if the system is off resonance , however , the system can be averaged with the result that @xmath65 averages out to zero and the binary system simply collapses due to gravitational radiation damping .
the average value of @xmath65 at the @xmath69 resonance is in general nonzero only for @xmath70 ; otherwise , the binary system simply collapses on average just as it does off resonance .
once the system is captured into a primary @xmath71 resonance , the resonance condition is not rigorously maintained .
there are in fact deviations that have general characteristics . to study these
, we let @xmath72 where @xmath73 and @xmath74 are new variables associated with resonance and @xmath75 is the corresponding small parameter .
the average energy exchange due to the external perturbation is generally oscillatory @xmath76 whereas the damping is unidirectional ; in fact , we expect on the average slow oscillatory behavior about the resonance manifold .
this motion can be described in terms of a general damped pendulum - like equation in @xmath73 with torque exerted by the radiation damping .
the equations of motion averaged around the @xmath71 resonance and expressed to second order in @xmath75 have been given in @xcite for the planar kepler system under radiative perturbations and in @xcite for the corresponding three - dimensional case . restricting the results of @xcite to the circularly - polarized incident wave under consideration here ( see equations 18 - 19 of @xcite ) , the second - order partially averaged equations at the @xmath71 resonance are @xmath77 \nonumber \\ & \quad-\epsilon { \cal d } \big [ -m\frac{\partial t_{c}}{\partial
l } \sin { m \phi } + m \frac{\partial t_{s}}{\partial l } \cos { m \phi } + \frac{\delta}{3l_{\star}^{3}g^{5}}(146 + 37e^{2 } ) \big ] , \nonumber \\ \dot{g } & = -\epsilon \big [ \frac{\partial t_{c}}{\partial \hat{g } } \cos { m \phi } + \frac{\partial t_{s}}{\partial \hat{g } } \sin { m \phi } + \frac{\delta}{l_{\star}^{3}g^{4}}(8 + 7e^{2 } ) \big ] .
\end{aligned}\ ] ] here @xmath78 , @xmath79 , and @xmath80 where @xmath81 .
the general behavior of the functions @xmath82 has been discussed in @xcite ( see figure [ fig1 ] and appendix a of @xcite ) .
the functions @xmath83 and @xmath84 are both negative for @xmath85 . in general , the derivative of the function @xmath86 is negative for @xmath87 ( see p. 107 of @xcite ) .
moreover , for @xmath88 , @xmath89 is positive for @xmath90 , zero for @xmath91 , and negative for @xmath92 ; in fact , @xmath93 , @xmath94 , etc . it is interesting to note that for incident radiation that is circularly polarized , equation implies that @xmath95 . introducing a new variable @xmath96 and a new slow temporal parameter @xmath97 , where @xmath98 and @xmath99 , system reduces to the form @xmath100 , \nonumber \\
\fl\dot{\cal d } = mb\hat{f } \sin{\theta } + \frac{\delta b}{g^{7 } } ( 8+\frac{73}{3}e^{2}+\frac{37}{12}e^{4 } ) + \mu b { \cal d } \big[m\hat{f}_{l } \sin{\theta } + \frac{\delta}{3l_{\star}^{3}g^{5}}(146 + 37e^{2 } ) \big ] , \nonumber \\
\fl\dot{g } = \mu b \big [ \pm 2\hat{f } \sin { \theta } + \frac{\delta}{l_{\star}^{3 } g^{4 } } ( 8 + 7e^{2 } ) \big ] , \end{aligned}\ ] ] where the _ overdot now refers to the new time _ @xmath101 and all relevant quantities are evaluated at @xmath102 .
for example , @xmath103 , etc .
it follows from equation that the first - order averaged equations reduce to @xmath104 , where @xmath105 is a constant , and @xmath106 , where @xmath107 is a solution of @xmath108 here @xmath109 and @xmath110 are constants given by @xmath111 where @xmath112 .
equation describes a mathematical pendulum with a constant torque @xmath113 ; the phase portrait of this dynamical system is illustrated in figure [ fig1 ] for @xmath114 . in this case
, there is a homoclinic orbit which encloses a region corresponding to orbits that are captured into resonance .
the region outside the homoclinic loop , with the stable and unstable manifolds removed , consists of orbits that pass through the resonance .
the presence of the perturbation terms in the system implies that the pendulum is damped ( or antidamped ) in the second - order averaged equations . that is , averaging the dynamical system to second order results in a pendulum equation of the form with slowly drifting @xmath109 and @xmath110 as well as an additional term proportional to @xmath115 .
the motion is geometrically described in the next section .
there are several important phenomena associated with resonance ; for instance , chaos and resonance capture .
chaotic motion is expected to occur near a resonance under a hamiltonian perturbation . for non - hamiltonian perturbations ,
while chaotic motions are certainly possible , resonance capture is often the dominant phenomenon . at first order ,
the partially averaged equations at a resonance are generally pendulum - like equations with torque .
this nonlinear hamiltonian system can exhibit chaos and resonance capture under perturbation .
the relationship between the chaos in the original dynamical system and possible chaotic effects in the averaged system near @xmath71 resonance is rather subtle and beyond the scope of this work .
here , we will consider instead the possibility of resonance capture . indeed
, we will give a geometric description of resonance capture in system and a rigorous proof that resonance capture does in fact occur .
we start from our model equations in delaunay variables and introduce the transformation about a primary resonance ; then , the resulting system can be transformed into the system plus terms of higher order by means of an averaging transformation .
these higher - order terms may be nonnegligible and dealing with this problem is beyond the scope of our work ; hence , we consider here resonance capture only in the second - order partially averaged equations and not in the full model .
system is typical of the following class of differential equations @xmath116 where @xmath117 and @xmath118 . in particular , note that @xmath119 is changing on a slow time scale relative to @xmath120 . for system ,
the pair @xmath121 plays the role of @xmath120 , and @xmath50 , in our case a scalar variable , plays the role of @xmath119 . for
the geometry of resonance capture , let us make a basic hyperbolicity assumption about system , namely , let us assume that for each fixed @xmath122 in some bounded region @xmath123 of @xmath124 the planar differential equation @xmath125 has a hyperbolic saddle point @xmath126 with a corresponding homoclinic loop given by a solution @xmath127 .
then , the set @xmath128 is a normally hyperbolic invariant manifold ( consisting entirely of rest points ) for the unperturbed system , i.e. @xmath129 a rough description of this manifold is that nearby orbits are attracted or repelled from the vicinity of @xmath130 exponentially fast and the rates of expansion and contraction dominate the corresponding rates for orbits on @xmath130 . of course , in our case
where the manifold consists of rest points , the second condition is satisfied automatically .
the normally hyperbolic manifold @xmath130 forms the `` spine '' of the submanifold @xmath131 in @xmath132 consisting of the manifold @xmath130given as a graph over an open subset of @xmath124and all the corresponding homoclinic loops as illustrated in figure [ fig2 ] .
the manifold @xmath131 `` bounds '' a region in the phase space for system , and this region corresponds to the initial conditions for orbits that are captured into resonance relative to system .
when the system is subjected to perturbation , @xmath131 is expected to `` split '' , for example , as depicted in figure [ fig3 ] .
in fact , by a celebrated theorem of n. fenichel @xcite , the normally hyperbolic manifold @xmath130 persists .
if the perturbation parameter @xmath133 is sufficiently small , then the perturbed system has an `` invariant '' normally hyperbolic manifold @xmath134 that is again given as a graph over a region in @xmath124 ; in particular , the manifold @xmath134 has stable and unstable manifolds .
under appropriate conditions that will be specified below , the stable and unstable manifolds of @xmath134 , the manifolds resulting from the split homoclinic manifold @xmath131 , form a channel in the phase space where orbits of the phase flow enter the vicinity of the resonance ( cf .
figure [ fig3 ] ) .
this is the geometric description of resonance capture for a system such as .
let us note that because the unperturbed `` slow manifold '' @xmath130 consists entirely of rest points , it is an invariant set for the unperturbed flow .
however , the perturbed manifold @xmath134 generally has no fixed points ; rather , the flow on @xmath134 is characterized by a slow drift in the @xmath119 variable .
thus , @xmath134 is only invariant in the sense that orbits with initial conditions on this manifold stay on it until they exit at one of its boundary points . of course
, as an effect of the slow drift , an orbit can , after a ( perhaps long ) sojourn near the resonance , enter a new dynamical regime .
the sojourn time can be predicted from an analysis of the flow on the slow manifold .
thus , our geometric picture gives a rather complete description of the phenomena associated with resonance capture . to prove that the geometric description of resonance capture obtains for system
, we must detect splitting of @xmath131 and determine the positions of the stable and unstable manifolds of @xmath134 .
this is accomplished by an adaptation of melnikov s method that was first carried out by robinson @xcite .
to state his result , let us choose a portion of a plane @xmath135 in @xmath132 that intersects each homoclinic loop in @xmath131 perpendicular to the unperturbed vector field on @xmath131 .
the choice of @xmath135 gives a smooth choice of initial conditions for the unperturbed orbits that correspond to the homoclinic loops .
indeed , for each @xmath122 , consider the corresponding saddle point @xmath126 and the unperturbed solution @xmath136 such that @xmath137 corresponds to the intersection point of @xmath135 with the homoclinic loop at @xmath126 .
let @xmath138 denote the distance between the intersections of the stable and unstable manifolds of @xmath134 as measured along @xmath135 in the intersecting plane parallel to the @xmath120 coordinate plane and passing through the point @xmath139 .
also , for two vectors @xmath140 , let @xmath141 .
the following theorem is proved in @xcite : [ robth ] if system satisfies the basic hyperbolicity assumption and the vector field given by @xmath142 is divergence - free for each fixed @xmath143 , then the splitting distance is given by @xmath144 , where the leading - order coefficient is given by the integral @xmath145 all functions in the integrand are evaluated at @xmath146 , and the function @xmath147 satisfies the variational initial value problem @xmath148 the following immediate corollary is also stated in @xcite : if @xmath149 for all @xmath143 , then there is resonance capture .
moreover , the only way a captured solution can leave the vicinity of the resonance is to have its second component , @xmath119 , leave the region where @xmath149 , a process that occurs on a timescale of order @xmath150 .
the existence of resonance capture means that a capture channel opens up as in figure [ fig3 ] and therefore there is sustained resonance .
let us now formulate and prove a version of another result of robinson @xcite that we use to apply theorem [ robth ] to the pendulum - type systems that arise by partial averaging at resonance .
recall that system is derived from the averaged physical model in part by a reversal of time .
thus , it is convenient to have a result that implies `` release from resonance '' for system so that it will imply capture into resonance for the original averaged model .
of course , the phrase `` release from resonance '' means that the stable and unstable manifolds of the normally hyperbolic manifold @xmath130 split in the direction that is opposite to that required for capture ; in effect , orbits leave the vicinity of the resonance in a channel bounded by the unstable manifold .
[ rt ] consider the system @xmath151 where @xmath152 , @xmath153 , and all indicated functions are smooth . if there is a point @xmath154 such that @xmath155 then the unperturbed system @xmath156 has a homoclinic loop that encloses a region @xmath157 in the @xmath158-plane .
suppose that , in addition , @xmath159 is a solution on the homoclinic loop with @xmath160 where a prime denotes differentiation with respect to @xmath161 , and for @xmath162 , @xmath163 if at least one of the inequalities , , or is strict , then robinson s integral @xmath164 in is negative at @xmath161 .
in particular , the perturbed configuration of stable and unstable manifolds corresponds to `` release from resonance '' .
the first statement of the proposition , about the existence of homoclinic loops , is easily proved by elementary phase - plane analysis .
let @xmath165 denote a new time - like variable given by @xmath166 such that @xmath167 at @xmath168 and let @xmath169 . in these new variables ,
system has the form @xmath170 robinson s integral @xmath171 is then given by @xmath172 where @xmath173 .
we find that @xmath174 , where @xmath175\,\rmd s , \\
i_2:=\frac{1}{2}\int^\infty_{-\infty } f^{-\frac{3}{2}}v^2f'r\,\rmd s , \\
i_3:=-\int_{-\infty}^\infty v\big(g / f\big)'{{\bi{w}}}_\mu\,\rmd s. \end{aligned}\ ] ] also , let us note that @xmath161 is constant on the separatrix loop and let us take the solution of system such that the partial derivative @xmath176 vanishes at @xmath167 as required in the statement of theorem [ robth ] .
the first integral @xmath177 can be rewritten as follows : @xmath178f^{1/2}\,\rmd t\\ & = \frac{1}{f}\int^\infty_{-\infty } [ p(f\sin{\varphi}+g)-vq]\,\rmd t. \end{aligned}\ ] ] under the hypothesis , the homoclinic loop @xmath179 in the @xmath180-plane is parametrized with negative orientation relative to the usual orientation of the plane .
also , let us denote by @xmath181 the bounded region with boundary @xmath179 in the @xmath180-plane . by green s theorem and the inequality
, we have that @xmath182 it follows from hypothesis that @xmath183 . for
the integral @xmath184 , let us observe that if we take the parametrization of @xmath179 such that @xmath185 , then @xmath186 ; that is , @xmath187 has the same sign as the variable @xmath165 in the integration .
also , because @xmath161 is constant for the integration , @xmath188 for the functions in the integrand , and hypothesis holds , we have the inequality @xmath189 = \big(\frac{g}{f}\big ) ' \frac{\rmd{{\bi{w}}}_\mu}{\rmd s}\big|_{\mu = 0 } = \big(\frac{g}{f}\big ) ' r f^{-1/2}\geq 0.\ ] ] thus , using the fact that @xmath176 vanishes at @xmath167 , note that @xmath190 has the same sign as the variable @xmath165 for the integration .
we conclude that @xmath191 over the range of the integration .
thus , @xmath192 .
finally , note that if inequality is strict , then @xmath193 .
likewise , if inequality ( respectively inequality ) is strict , then @xmath194 ( respectively @xmath195 ) .
hence , @xmath196 and if at least one of the inequalities , , or is strict , then @xmath197 . to apply proposition [ rt ] ,
let us consider the second - order averaged equations for the normally incident left circularly polarized gravitational wave expressed in decoupled form ; namely , system with the `` lower '' signs . also , let us note that this system is a special case of the abstract system in proposition [ rt ] once we set @xmath198 , @xmath199 , and @xmath200 . recall that if @xmath201 , then @xmath85 and @xmath202 .
hence , we have the desired inequality @xmath203 once we choose @xmath204 .
in fact , we set @xmath205 .
we claim that the hypotheses of proposition [ rt ] are satisfied under the following two conditions : 1 .
there is a subinterval @xmath206 of the interval @xmath207 such that @xmath208 whenever @xmath209 .
2 . there is an interval @xmath210 such that the derivative of the function @xmath211 is negative on the image of the interval @xmath212 under the transformation @xmath213 condition 1 ensures that hypothesis is satisfied for each @xmath209 .
recall that @xmath214 and since @xmath215 for @xmath87 we have that @xmath216 .
thus , @xmath217 for @xmath209 .
turning to the function @xmath218 in the statement of proposition [ rt ] , note that @xmath219 by the first implication of proposition [ rt ] , it follows that if @xmath209 , then the system @xmath220 has a homoclinic orbit where the corresponding saddle point has coordinates @xmath221 for some @xmath222 in the interval @xmath223 as long as @xmath224 . at the boundary of the interval @xmath206 , where @xmath225 the corresponding saddle point with coordinates @xmath226 and center with coordinates @xmath227 coalesce at @xmath228 ; that is , these rest points disappear in a saddle node bifurcation .
thus , there is a subinterval @xmath210 such that if @xmath229 , then the region @xmath230 , enclosed by the homoclinic orbit , contains only points whose first coordinates @xmath231 are in the interval @xmath232 . in particular , if @xmath229 , then @xmath233 and therefore @xmath234 .
hence , @xmath235 and hypothesis is satisfied . to verify hypothesis
, we must show that if @xmath229 , then the derivative of the function @xmath236 is positive ; however , this requirement follows immediately from condition 2 . for hypothesis ,
let us determine the sign of the divergence of the vector field @xmath237 given by @xmath238.\ ] ] indeed , using the inequalities @xmath233 and @xmath217 , it follows that if @xmath229 , then the divergence of @xmath237 is the positive quantity @xmath239 this completes the proof of our claim . to show the desired phenomenon of release from resonance
, we will verify conditions 1 and 2 . in order to determine the interval @xmath212 ,
let us note that our ( unperturbed ) pendulum - type planar system has the form @xmath240 where @xmath241 , @xmath242 , and @xmath243 .
system is hamiltonian with total energy @xmath244 denote the coordinates of its hyperbolic saddle point by @xmath226 , and recall that @xmath245 .
also , note that the homoclinic loop is given by the graphs of the functions @xmath246 [ p1 ] for system , there is a point @xmath247 such that @xmath248 and @xmath249 the point @xmath250 is not a rest point .
moreover , the equation @xmath251 has no solution for @xmath252 .
that is , the homoclinic loop at @xmath253 crosses the @xmath254-axis at @xmath255 and encloses that portion of the @xmath254-axis given by @xmath256 .
also , @xmath257 if and only if @xmath258 is such that @xmath259 define the function @xmath260 by @xmath261 and note that it has relative maxima at @xmath262 and @xmath263 .
also , @xmath264 and @xmath265 .
it follows that there is a point @xmath247 in the interval indicated in the statement of the proposition such that @xmath266 .
moreover , this point is not a rest point .
indeed , the only rest point in the open interval @xmath267 is the point @xmath268 , and it corresponds to a negative relative minimum of @xmath260 .
if @xmath269 , and @xmath270 , then there must be some relative maximum of @xmath260 that is nonnegative .
however , all the relative maxima that exceed @xmath222 have the form @xmath271 , where @xmath272 is a positive integer , and @xmath260 takes negative values at all such points ; therefore , we reach a contradiction . clearly , @xmath257 if and only if @xmath273 or @xmath274 recall that @xmath275 and @xmath276
thus , with @xmath258 , inequality is equivalent to the inequality @xmath277 as required . in summary ,
let us recall that @xmath278 , where @xmath279 and @xmath280 by equation .
for incident left circularly polarized radiation , we have that @xmath281 and @xmath282 ; in fact , @xmath283 and for @xmath284 the first term in this taylor expansion is accurate to about 10% .
then , all the hypotheses of proposition [ rt ] are satisfied provided that there is some @xmath43 such that @xmath85 and @xmath285 one can show by inspection that condition ( iii ) is satisfied provided that @xmath286 , where @xmath287 . we therefore choose @xmath288 in agreement with condition ( i ) .
moreover , one can directly verify that @xmath289 if we assume @xmath290 and @xmath291 .
it is interesting to note that the eccentricities of the relativistic binary pulsars psr b1913 + 16 and psr b1534 + 12 are @xmath292 and @xmath293 , respectively .
we choose @xmath294 , so that @xmath295 at the ( 2 : 1 ) resonance that we wish to consider .
finally , @xmath60 is uniquely determined from @xmath296 once we recall that @xmath205 and @xmath297 ; hence , @xmath298 .
we have shown that a set of parameters exists such that for incident left circularly polarized radiation the averaged equations permit sustained resonance .
the essential requirement for sustained resonance in our approach is that the robinson integral be nonzero .
our objective has been to _ prove _ sustained resonance in the second - order averaged equations for our model .
however , we expect that robinson s integral is generally nonzero without the restrictive assumptions of proposition [ rt ]
. indeed , it should be clear from the proof of proposition [ rt ] that @xmath171 can be negative without the requirement that each of the three integrals @xmath177 , @xmath299 , and @xmath184 be negative . for this reason , we expect that resonance capture would generally be possible for either circular polarization state of the incident radiation .
moreover , circular polarization is not essential for sustained resonance as demonstrated by our previous numerical work on resonance capture that did not involve circular polarization [ 2,4,5 ] .
indeed , the mathematical theory of sustained resonance developed in this paper owes much to the simplification
the reduction from four to three phase - space dimensions in the system brought about by the assumption that the normally incident radiation has definite helicity .
the proof of the existence of sustained resonance involves the second - order partially averaged system .
we conjecture that the result can be extended to the original system .
to provide evidence for this conjecture , it is interesting to use the set of parameters obtained in the previous section as part of the initial conditions needed to integrate the original system numerically . to this end , we integrate system with initial conditions @xmath300 , @xmath301 , @xmath302 , and @xmath303 .
thus the initial orbit has @xmath304 , @xmath290 , @xmath305 , and @xmath306 , so that @xmath307 as required .
the periastron for this initial osculating orbit , which has true anomaly @xmath308 , occurs at @xmath309 ; hence @xmath310 .
in fact , this initial orbit has been used frequently in our previous work [ 36 ] .
we find that with these initial conditions the system is indeed in sustained ( 2 : 1 ) resonance for either circular polarization state as demonstrated in figures 4 and 5 .
figure [ fig4 ] can be produced as follows : system is integrated backward from the initial point for normally incident radiation that is left circularly polarized until the system leaves the resonance and a final point of integration is stored ; then , the graph is made by integrating forward in time from this point .
it is interesting to note that the binary orbit collapses rather rapidly following exit from resonance when the orbit is highly eccentric .
this is explained by the fact that the rate of emission of gravitational radiation by the binary orbit is proportional to @xmath311 ; hence , as @xmath312 the loss of orbital energy to radiation leads to orbital collapse @xcite .
figure [ fig5 ] can be produced in a similar way and illustrates the same situation for right circular polarization .
the _ possibility _ of resonance in this case can be seen from the first - order averaged equations ; that is , @xmath313 and hence @xmath314 , so that the situation depicted in figure 1 holds except that the direction of some arrows should be reversed .
it is clear from figures 4 and 5 that evolutionary dynamics of the orbit while trapped in resonance depends sensitively upon the helicity of the incident radiation . for example
, the orbit loses angular momentum while in resonance and becomes highly eccentric when the incident radiation has negative helicity as in figure 4 , while the opposite situation holds in figure 5 for positive helicity incident radiation ; this circumstance may be naturally interpreted in terms of the ( negative or positive ) angular momentum carried by the ( negative or positive helicity ) radiation and deposited into the binary orbit .
the fact that the amplitude of oscillations around resonance is generally larger when the sense of the helicity of the radiation is the same as the rotation of the orbit is in agreement with our previous work on the hill system . in contrast to the hill system , the present case involves dissipation ; nevertheless , the analysis of the amplitude of oscillations around the resonance is very similar to that carried out in section 6 of @xcite due to the fact that @xmath315 .
figures 4 and 5 also illustrate the complex structure that is usually associated with a higher - order resonance @xcite ; furthermore , the presence of chaos is strongly indicated in our numerical work .
chaos is especially noteworthy in figure 5 ; therefore , the qualitative significance of our numerical results should be emphasized .
we have performed numerical experiments with @xmath0 and @xmath33 twice as large as those in figures 4 and 5 . in this case , we find in the experiment corresponding to figure 5 that the @xmath316 resonance appears to have significant overlap with the @xmath317 resonance .
let us note that we expect @xmath0 to be many orders of magnitude smaller in physical situations of interest ; however , we have taken @xmath0 to be @xmath318 in figures 4 and 5 for the sake of simplicity .
do all relativistic binary orbits _ monotonically _ decay by gravitational radiation damping ?
there is evidence for such energy loss from the timing analysis of only two pulsar systems @xcite .
however , such a system could in principle fall into sustained resonance with an external source ; during resonance lock , the orbit would not decay on the average .
the orbital decay would resume once the system is released from resonance .
it remains to see as more binary pulsars are discovered and their timing data accumulate whether sustained resonance in fact occurs in relativistic binary systems .
carmen chicone was supported by the nsf grant dms-9531811 and the university of missouri research board .
99 chicone c , mashhoon b and retzloff d g 1996 gravitational ionization : periodic orbits of binary systems perturbed by gravitational radiation _ ann . inst .
h. poincar , phys .
thor . _ * 64 * 87125 chicone c , mashhoon b and retzloff d g 1996 on the ionization of a keplerian binary system by periodic gravitational radiation _ j. math .
phys . _ * 37 * 39974016 ; 1997 addendum _ ibid . _ * 38 * 544 chicone c , mashhoon b and retzloff d g 1997 gravitational ionization : a chaotic net in the kepler system _ class .
quantum grav . _
* 14 * 699723 chicone c , mashhoon b and retzloff d g 1997 evolutionary dynamics while trapped in resonance : a keplerian binary system perturbed by gravitational radiation _ class .
quantum grav . _
* 14 * 18311850 chicone c , mashhoon b and retzloff d g 1999 chaos in the kepler system _ class .
quantum grav . _
* 16 * 507527 chicone c , mashhoon b and retzloff d g 1999 chaos in the hill system _ helv . phys .
acta _ , in press henrard
j 1985 resonance sweeping in the solar system _ stability of the solar system and its minor natural and artificial bodies _ ed v g szebehely ( dordrecht : reidel ) pp 183192 beaug c , aarseth s j and ferraz - mello s 1994 resonance capture and the formation of the outer planets _ mon . not .
r. astron .
soc . _ * 270 * 2134 melita m d and woolfson m m 1996 planetary commensurabilities driven by accretion and dynamical friction _ mon . not .
r. astron .
soc . _ * 280 * 854862 winter o c and murray c d 1997 resonance and chaos .
i. first - order interior resonances _ astron .
astrophys . _ * 319 * 290304 haghighipour n 1999 dynamical friction and resonance trapping in planetary systems _ mon . not .
r. astron .
_ * 304 * 185194 haghighipour n 1999 resonance lock and planetary dynamics _ ph.d .
thesis _ university of missouri - columbia wardell z e 1999 gravitational radiation reaction and the three body problem , in preparation kopeikin s m 1998 , private communication kopeikin s m 1997 millisecond and binary pulsars as nature s frequency standards .
i. a generalized statistical model of low - frequency timing noise _ mon . not .
r. astron .
* 288 * 129137 kopeikin s m 1999 millisecond and binary pulsars as nature s frequency standards .
ii . effects of low - frequency timing noise on residuals and measured parameters _ mon . not .
r. astron .
_ * 305 * 563590 bombelli l and calzetta e 1992 chaos around a black hole _ class .
quantum grav . _
* 9 * 25732599 stairs i h , arzoumanian z , camilo f , lyne a g , nice d j , taylor j h , thorsett s e and wolszczan a 1998 measurement of relativistic orbital decay in the psr b1534 + 12 binary system _ astrophys .
j. _ * 505 * 352357 taylor j h and weisberg j m 1989 _ astrophys .
j. _ * 345 * 434450 kovalevsky j 1967 _ introduction to celestial mechanics _
( _ astrophysics and space science library _ vol 7 ) ( new york : springer ) sternberg s 1969 _ celestial mechanics _
vols 12 ( new york : benjamin ) fenichel n 1971 persistence and smoothness of invariant manifolds for flows _ indiana univ .
j. _ * 21 * 193226 robinson c 1983 sustained resonance for a nonlinear system with slowly varying coefficients _ siam j. math .
* 14 * 847860 | the general phenomena associated with sustained resonance are studied in this paper in connection with relativistic binary pulsars .
we represent such a system by two point masses in a keplerian binary system that evolves via gravitational radiation damping as well as an external tidal perturbation . for further simplification ,
we assume that the external tidal perturbation is caused by a normally incident circularly polarized monochromatic gravitational wave . in this case , the second - order partially averaged equations are studied and a theorem of c. robinson is employed to prove that for certain values of the physical parameters resonance capture followed by sustained resonance is possible in the averaged system .
we conjecture that sustained resonance can occur in the physical system when the perturbing influences nearly balance each other . |
a recent progress in a targeted design of molecular nanomagnets , which afford paradigmatic examples of assemblies of a finite number of interacting spin centers , has triggered a considerable interest to explore a magnetic behavior of small spin clusters magnetically isolated from the environment @xcite . despite their simplicity , small spin clusters may still exhibit favorable magnetic properties such as the quantum tunneling of magnetization and slow spin relaxation observable in single - molecule magnets , which have an immense application potential in a development of new resources suitable for quantum information processing and/or novel generation of high - density storage devices @xcite .
in addition , the molecular nanomagnets being composed of small spin clusters are ideal for an experimental testing of the limitations of physical theories especially when physical properties of relevant model systems can be calculated without any uncontrolled approximation .
magnetic properties of small spin clusters are predominantly determined by the nature of magnetic interactions between the spin centers in a cluster , whereas the heisenberg superexchange coupling is usually the most dominant interaction term that basically influences characteristic features of molecular nanomagnets @xcite .
a lot of attention has been therefore paid to the study of antiferromagnetic heisenberg spin clusters , which may exhibit striking intermediate plateaux in low - temperature magnetization curves that often macroscopically manifest highly non - trivial quantum ground states @xcite .
an influence of the exchange , dipolar and single - ion anisotropy on the low - temperature magnetization process of the antiferromagnetic heisenberg spin clusters has attracted much less attention so far @xcite .
for this reason , it appears worthwhile to investigate the low - temperature magnetization process of antiferromagnetic ising spin clusters , which are also capable of displaying several intermediate magnetization plateaux on assumption that a spin cluster is geometrically frustrated @xcite .
the main goal of the present work is to examine the magnetization process and adiabatic demagnetization of the antiferromagnetic ising spin clusters with the shape of regular polyhedra ( platonic solids ) , which surprisingly seem not to be dealt with previously .
the zero - field thermodynamics of the regular ising polyhedra with the uniform interaction have been explored in some detail by syozi @xcite and fisher @xcite within the framework of dual , decoration - iteration and star - triangle transformations , while the regular ising polyhedra with the mixed ferromagnetic and antiferromagnetic couplings have been examined in relation with the spin - glass physics @xcite .
it is noteworthy that the competition between the antiferromagnetic order , spin frustration and magnetic field has been studied by viitala and co - workers by considering antiferromagnetic ising spin clusters with several cluster geometries as for instance octahedron , body centered icosahedron and cubooctahedron @xcite .
more recently , an exact enumeration of states has been employed in order to calculate numerically the residual entropy , magnetocaloric properties and magnetization process of various geometrically frustrated ising spin clusters composed of triangular units @xcite .
the organization of this paper is follows .
exact results for the partition function , free energy and magnetization of the regular ising polyhedra are derived in section [ method ] within the framework of a graph - theoretical approach .
the most interesting results obtained for the magnetization process and adiabatic demagnetization of the regular ising polyhedra are presented and discussed in detail in section [ result ] .
finally , some conclusions and future outlooks are drawn in section [ conclusion ] .
let us consider the ising spin clusters with the geometry of five regular polyhedra ( tetrahedron , octahedron , cube , icosahedron and dodecahedron ) as depicted in fig .
the regular ising polyhedra can be defined through the following hamiltonian @xmath0 where @xmath1 represents the ising spin placed at @xmath2th vertex of a regular polyhedron , the first summation accounts for the antiferromagnetic ising - type exchange interaction @xmath3 between adjacent spins , the second summation accounts for the zeeman s energy of individual magnetic moments in the external magnetic field @xmath4 and finally , @xmath5 @xmath6 stands for the total number of vertices ( edges ) of a regular polyhedron that is simultaneously equal to the total number of spins ( bonds ) .
it is noteworthy that the considered ising spin clusters are sufficiently small finite - size systems in order to calculate all physical quantities of interest quite rigorously . in the following , we will calculate the partition function , free energy and magnetization of the regular ising polyhedra within a relatively simple graph - theoretical approach .
the canonical partition function of the regular ising polyhedra is defined through the relation @xmath7 where @xmath8 , @xmath9 is boltzmann s constant , @xmath10 is the absolute temperature and the summation @xmath11 is carried out over all possible configurations of a set of the ising spins @xmath12 . while the interaction part of the hamiltonian ( the first term in eq .
( [ ham ] ) ) depends on many specific details of a particular spin configuration , the zeeman s energy ( the second term in eq .
( [ ham ] ) ) solely depends on the total spin @xmath13 and is independent of any other specific detail of a given spin configuration .
it is therefore quite convenient to sort available spin configurations according to the total spin @xmath14 , because only the lowest - energy spin configuration from a given set of microstates with the same total spin @xmath14 may eventually become a ground state .
in addition , it is sufficient to consider only spin configurations with non - negative values of the total spin @xmath15 , since the exchange energy is invariant under the reversal of all ising spins ( @xmath16 for each @xmath2 ) and the zeeman s energy merely changes its sign . to proceed further with a calculation ,
let us exploit the one - to - one correspondence between the ising spins @xmath1 and the newly defined two - valued variables @xmath17 established through the formula @xmath18 which connects the up spin state @xmath19 to the value @xmath20 and the down spin state @xmath21 to the other value @xmath22 .
the hamiltonian ( [ ham ] ) can be consequently rewritten in terms of newly defined two - valued variables @xmath23 to the following form @xmath24 \nonumber \\ \!\!\!&-&\!\!\ !
h \left(n - 2 \sum_{i=1}^{n } n_i \right ) . \label{hamn}\end{aligned}\ ] ] it is quite evident from eq .
( [ hamn ] ) that the exchange energy @xmath25 of the fully aligned spin configuration can be lowered by an amount @xmath26 per each unequally oriented pair of adjacent ising spins , but each ising spin flipped in opposite to the external magnetic field @xmath21 @xmath27 simultaneously raises the zeeman s energy @xmath28 by an amount @xmath29 .
apparently , the lowest - energy spin configuration ( ground state ) at a given magnetic field must be a reasonable compromise that would minimize a sum of both parts of the overall energy .
the alternative form of the hamiltonian ( [ hamn ] ) also suggest an interesting mapping correspondence between the summation over all available configurations of the ising spins and a relatively simple graph - theoretical counting problem .
when the ising spins oriented in opposite to the magnetic field @xmath30 ( @xmath31 ) are identified as vertices of the equivalent planar graph ( see lower panel in fig .
[ fig1 ] ) , the combinatorial problem of finding all possible spin configurations becomes equivalent to the problem of finding all induced subgraphs of a given planar graph . as a matter of fact , the total number of the flipped ising spins ( @xmath30 , @xmath31 ) then directly equals to the total number of vertices @xmath32 of the corresponding induced subgraph and , respectively , the total number of unequally oriented adjacent spin pairs @xmath33 equals to a sum over complementary degrees of all vertices present in a given induced subgraph @xmath34 ( @xmath35 is degree of a vertex in a full graph and @xmath36 is degree of the @xmath2th vertex in a given induced subgraph ) .
hence , it follows that the overall energy of each individual spin configuration can be straightforwardly calculated from the total number of vertices @xmath37 and the sum of all complementary degrees @xmath38 of the corresponding induced subgraph according to the formula @xmath39 it is worthwhile to remark that the graphs corresponding to the regular ising polyhedra are vertex - transitive , which substantially simplifies the relevant calculation of complementary degrees of vertices in respective induced subgraphs .
let us illustrate the aforedescribed calculation procedure on particular case of the regular ising polyhedra starting from the simplest example , which is the ising tetrahedron ( see tab .
when all four spins are in the up spin state @xmath40 ( @xmath41 ) one obtains a fully aligned spin configuration of the ising tetrahedron with the total spin @xmath42 , which corresponds according to the established mapping equivalence to a very special subgraph @xmath43 ( empty graph ) with a zero vertex set .
all other microstates of the ising tetrahedron with @xmath15 can be then obtained from this fully aligned spin configuration by turning a desired number of the ising spins to their down spin state , whereas it is sufficient to flip no more than a half of the total number of spins owing to the invariance of the exchange energy with respect to the inter - change @xmath44 .
bearing this in mind , only two additional spin configurations of the ising tetrahedron should be taken into account with either one or two inverted spins , whereas the former spin configuration with @xmath45 corresponds to an induced subgraph containing an isolated vertex ( the subgraph 1a in fig .
[ fig2 ] ) and the latter spin configuration with @xmath46 to an induced subgraph involving two vertices incident to a common edge ( the subgraph 2b in fig .
[ fig2 ] ) .
it is worthwhile to remember that the overall energy of spin configurations with @xmath47 can be obtained from that ones with @xmath48 simply by inverting the sign of zeeman s term ( the exchange energy is invariant under this reversal ) .
.spin configurations of the ising tetrahedron classified according to the total spin @xmath49 , the total number of flipped ising spins ( @xmath37 ) , the degeneracy ( deg ) , the total number of unequally oriented adjacent spin pairs ( @xmath38 ) , the overall energy and the corresponding induced subgraph ( see fig . [ fig2 ] for schematic representation of induced subgraphs ) .
the values @xmath37 and @xmath38 coincide with the total number of vertices in a given subgraph and the sum of their complementary degrees , respectively .
[ cols="^,^,^,^,^,^",options="header " , ] [ td ] at this stage , the overall energy spectrum of the regular ising polyhedra obtained within the simple graph - theoretical approach can be substituted to the definition of the partition function ( [ pf ] ) in order to get full thermodynamics . for completeness ,
let us merely quote the final exact expressions for the partition function of the ising tetrahedron @xmath50 the ising octahedron @xmath51 \nonumber \\
12\exp(-4\beta j)\cosh(4\beta h ) \nonumber \\ \!\!\!&+&\!\!\
! 2\exp(-12\beta j)\cosh(6\beta h ) , \label{pfokta}\end{aligned}\ ] ] the ising cube @xmath52 \nonumber \\ \!\!\!&+&\!\!\ !
8\cosh(4\beta h)[4 + 3\exp(-4\beta j ) ] \nonumber \\ \!\!\!&+&\!\!\ !
16\exp\left(-6\beta j\right)\cosh(6\beta h ) \nonumber \\ \!\!\!&+&\!\!\
! 2\exp(-12\beta j)\cosh(8\beta h ) , \label{pfhexa}\end{aligned}\ ] ] and the ising icosahedron @xmath53 \nonumber \\
120\cosh(2\beta h ) [ \exp(-4\beta j ) + \exp(-8\beta j ) ] \nonumber
\\ \!\!\!&+&\!\!\ ! 30\cosh(4\beta
h ) [ 2\exp\left(6\beta j\right ) + 13\exp\left(2\beta j\right ) ] \nonumber \\
120\cosh(4\beta h ) [ 3\exp\left(-2\beta j\right ) + \exp\left(-6\beta j\right ) ] \nonumber \\ \!\!\!&+&\!\!\ ! 60\exp\left(-10\beta j\right)\cosh(4\beta h ) \nonumber
\\ \!\!\!&+&\!\!\ !
40\exp(-12\beta j)\cosh(6\beta h ) \nonumber \\
40\cosh(6\beta h ) [ 1 + 6 \exp(-4\beta j ) + 3 \exp(-8\beta j ) ] \nonumber \\
12\cosh(8\beta h ) [ 6 \exp\left(-10\beta j\right ) + 5\exp\left(-14\beta j\right ) ] \nonumber \\
24\exp(-20\beta j)\cosh(10\beta h ) \nonumber \\ \!\!\!&+&\!\!\ !
2\exp\left(-30\beta j\right)\cosh(12\beta h ) . \label{pfikosa}\end{aligned}\ ] ] gibbs free energy normalized per one spin can be directly calculated from the partition functions of the regular ising polyhedra according to the relation @xmath54 which subsequently allows a straightforward calculation of the magnetization per spin @xmath55 as well as the entropy per spin @xmath56 for brevity , let us merely quote the final exact analytical formula for the single - site magnetization of the ising tetrahedron @xmath57 , \label{mtetra}\end{aligned}\ ] ] the ising octahedron @xmath58 \right . \label{mokta } \\ \!\!\!&+&\!\!\ ! \left .
4\exp(-4\beta j)\sinh(4\beta h ) + \exp(-12\beta j)\sinh(6\beta h ) \right\ } , \nonumber \end{aligned}\ ] ] the ising cube @xmath59 \right . \nonumber \\
2\sinh(4\beta h ) [ 3\exp(-4\beta j ) + 4 ] \label{mhexa } \\ \!\!\!&+&\!\!\ ! \left .
6\exp\left(-6\beta j\right)\sinh(6\beta h ) + \exp(-12\beta j)\sinh(8\beta h ) \right\ } , \nonumber \end{aligned}\ ] ] and the ising icosahedron @xmath60 \right .
\nonumber \\
10\sinh(2\beta h ) [ \exp(-4\beta j ) + \exp(-8\beta j ) ] \nonumber \\ \!\!\!&+&\!\!\ !
5\sinh(4\beta h ) [ 2\exp\left(6\beta j\right ) + 13\exp\left(2\beta j\right ) ] \nonumber
\\ \!\!\!&+&\!\!\ !
20\sinh(4\beta h ) [ 3\exp\left(-2\beta j\right ) + \exp\left(-6\beta j\right ) ] \nonumber \\ \!\!\!&+&\!\!\ ! 10\exp\left(-10\beta j\right)\sinh(4\beta h ) \nonumber \\
10\exp(-12\beta j)\sinh(6\beta h ) \nonumber \\ \!\!\!&+&\!\!\ !
10\sinh(6\beta h ) [ 1 + 6 \exp(-4\beta j ) + 3 \exp(-8\beta j ) ] \nonumber \\
4\sinh(8\beta h ) [ 6 \exp\left(-10\beta j\right ) + 5\exp\left(-14\beta j\right ) ] \nonumber \\ \!\!\!&+&\!\!\ !
10\exp(-20\beta j)\sinh(10\beta h ) \nonumber \\ \!\!\!&+&\!\!\ ! \left .
\exp\left(-30\beta j\right)\sinh(12\beta h ) \right\}. \label{mikosa}\end{aligned}\ ] ] it should be noted that the final expressions for the entropy of the regular ising polyhedra obtained according to eq .
( [ s ] ) are too cumbersome to write them down here explicitly .
in this part , let us explore in detail the magnetization process and magnetocaloric properties of the regular ising polyhedra .
in particular , we will investigate a microscopic nature of highly degenerate ground states , which are manifested in low - temperature curves as profound intermediate plateaux at fractional values of the saturation magnetization .
the magnetization of the ising tetrahedron is plotted in the upper panel of fig .
[ tetr ] against temperature and magnetic field .
it can be clearly seen from this figure that a zero - temperature magnetization curve of the ising tetrahedron shows two intermediate plateaux at zero and one - half of the saturation magnetization .
the magnetization is zero in a field range @xmath61 , which favors the six - fold degenerate ground state with two spins oriented in a direction of the magnetic field and the other two spins pointing in opposite direction .
one out of two spins oriented contrary to the magnetic field is flipped at the first critical field @xmath62 , above which the four - fold degenerate ground state emerges with three spins aligned into the magnetic field and one spin pointing in opposite direction .
the aforementioned ground state manifest itself as an intermediate plateau at one - half of the saturation magnetization , which can be detected in a field range @xmath63 .
finally , the fully polarized state with all four spins aligned towards the magnetic field becomes the ground state above the second critical field @xmath64 .
it should be pointed out that exact magnetization jumps and plateaux occur only at zero temperature , because a rising temperature generally causes a gradual smoothening of the magnetization curve . as a result ,
magnetization plateaux generally shrink and magnetization jumps are smeared out with increasing temperature .
let us also examine magnetocaloric properties of the ising tetrahedron , which can be particularly interesting especially in a vicinity of the aforedescribed magnetization jumps .
for this purpose , the lower panel of fig .
[ tetr ] displays a few isentropy lines of the ising tetrahedron in the magnetic field - temperature plane , which can be alternatively viewed as a temperature response with respect to a variation of the magnetic field under the adiabatic condition .
it actually turns out that the ising tetrahedron may exhibit an enhanced magnetocaloric effect in a vicinity of two magnetization jumps , above ( below ) which an abrupt rise ( drop ) in temperature is achieved by decreasing magnetic field whenever the entropy per spin is set close enough to the ground - state degeneracy pertinent to the first and second critical field @xmath65 and @xmath66 , respectively .
in contrast to the ising tetrahedron , the magnetization curve of the ising octahedron shows just one wide plateau at one - third of the saturation magnetization in a respective field range @xmath67 , see the upper panel in fig .
it is worthwhile to remark that the intermediate plateau at zero magnetization is totally absent in a relevant magnetization process , because three spin configurations with two antipodal spins oriented against the magnetic field are energetically favorable before any spin configuration with the total spin @xmath46 as soon as the magnetic field becomes non - zero . in virtue of this fact
, the zero - temperature magnetization curve of the ising octahedron exhibits an abrupt jump from zero to one - third of the saturation magnetization right at zero field , whereas the intermediate one - third plateau terminates at the saturation field @xmath68 that determines a field - induced transition towards the fully saturated state .
the effect of temperature upon a gradual smoothening of the magnetization curve is quite similar as in the case of the ising tetrahedron .
adiabatic changes of temperature achieved upon varying the magnetic field are depicted for the ising octahedron in the lower panel of fig .
apparently , the ising octahedron shows an enhanced magnetocaloric effect near the saturation field provided that the entropy per spin is close to the ground - state degeneracy relevant for the saturation field @xmath69
. however , it is worth noticing that an absence of intermediate plateau at zero magnetization has significant impact on magnetocaloric properties of the ising octahedron .
the absence of zero magnetization plateau bears a close relation to a giant magnetocaloric effect observable near zero magnetic field .
if the entropy is selected close enough to the ground - state degeneracy emergent at zero field @xmath70 , then , the adiabatic demagnetization of the ising octahedron is characterized by a rapid decrease in temperature as the magnetic field diminishes .
it should be stressed that the isentropy line at @xmath70 reaches the absolute zero temperature with an infinite gradient as the magnetic field tends to zero , which suggest the fastest possible cooling process that would be quite superior with respect to that of paramagnetic salts .
a refrigeration with the help of adiabatic demagnetization of the ising octahedron might be therefore of substantial technological importance , since the efficient cooling persists in a rather wide range of the magnetic fields and the ultra - low temperatures can be consequently reached quite efficiently .
the magnetization curve of the ising cube as the only non - frustrated regular polyhedron consists of a single plateau at zero magnetization , which is subsequently followed by a direct magnetization jump from zero to the saturation magnetization ( see the upper panel in fig . [ cube ] ) .
the bipartite nature of the ising cube enables an existence of two - fold degenerate antiferromagnetic ground state at sufficiently low magnetic fields with four reversed spins , which undergo a simultaneous spin flip to the fully polarized state at the saturation field @xmath64 .
it is noteworthy that the magnetization curve becomes smoother upon increasing temperature and there is a huge degeneracy precisely at the saturation field , at which all lowest - energy spin configurations with @xmath15 coexist together .
it could be concluded that the vertex - transitive bipartite ising spin clusters can not exhibit any other magnetization plateau except that at zero magnetization .
the adiabatic demagnetization of the ising cube is displayed in the lower panel of fig .
[ cube ] , which shows a contour plot of entropy as a function of temperature and magnetic field .
the giant magnetocaloric effect is observed in a vicinity of the saturation field on account of a huge degeneracy of the lowest - energy spin configurations with @xmath15 on assumption that the entropy per spin is set to @xmath71 .
however , a rather steep decrease in temperature observable slightly above the saturation field is successively followed by a subsequent steep increase in temperature below the saturation field , which consequently precludes a possible utilization of a giant magnetocaloric effect for cooling purposes .
the most remarkable magnetization process evidently exhibits the ising icosahedron , which displays in total three different intermediate plateaux at zero , one - third and one - half of the saturation magnetization ( see the upper panel in fig .
[ icos ] ) . at sufficiently low magnetic fields @xmath61 the ground state is highly ( 72-fold ) degenerate due to an accidental degeneracy of two lowest - energy spin configurations , in which six reversed spins are located at vertices that either form an open chain of six sites or a closed chain of five sites plus one additional isolated site ( see the corresponding induced subgraphs 6b and 6c in fig .
[ fig2 ] ) .
the intermediate one - third magnetization plateau develops at moderate fields @xmath72 on account of the another highly ( 30-fold ) degenerate ground state , in which two out of four reversed spins are located at isolated vertices and the other two inverted spins at adjacent vertices ( see the corresponding induced subgraph 4b in fig .
[ fig2 ] ) .
the last intermediate plateau at one - half of the saturation magnetization occurs in the field range @xmath73 , which prefers the highly ( 20-fold ) degenerate ground state with three isolated spins oriented in opposite to the magnetic field ( see the corresponding induced subgraph 3a in fig .
[ fig2 ] ) . apparently , the high degeneracy of all aforedescribed ground states is closely connected to a geometric spin frustration of the ising icosahedron , which is progressively lowered with an increase of the external magnetic field until the degeneracy is completely lifted above the saturation field @xmath74 due to the unique fully ordered spin configuration . to achieve an enhanced magnetocaloric effect during the adiabatic demagnetization of the ising icosahedron ( the lower panel in fig .
[ icos ] ) , it is necessary to set entropy close enough to an appropriate ground - state degeneracy at a critical field related to the respective magnetization jump .
thus , a gradual decrease in the magnetic field is responsible for the fast cooling ( heating ) slightly above ( below ) the relevant critical field only if entropy per spin is close to @xmath75 nearby the first critical field @xmath62 , @xmath76 nearby the second critical field @xmath64 and @xmath77 nearby the saturation field @xmath74 . the magnetization process of the ising dodecahedron ( the upper panel in fig .
[ dode ] ) is quite reminiscent of that of the ising octahedron .
the zero - temperature magnetization curve of the ising dodecahedron shows an abrupt jump at zero magnetic field , which is consecutively followed by a rather wide intermediate plateau at one - fifth of the saturation magnetization ending up only at the saturation field @xmath64 .
the ground state of the ising dodecahedron is just five - fold degenerate despite of a strong spin frustration in a low - field region @xmath78 , where five different spin configurations with eight isolated spins aligned contrary to the magnetic field become the lowest - energy microstates .
the magnetization plateau and jump are gradually smoothened with an increase in temperature until they completely vanish above a certain temperature .
one of the most important features of the magnetization process of the ising dodecahedron is the absence of zero magnetization plateau , which results from an incapability of microstates with the total spin @xmath79 or @xmath80 to further reduce the configurational energy of the ground - state manifold composed of five microstates with the total spin @xmath45 . owing to this fact ,
the giant magnetocaloric effect can be detected during the adiabatic demagnetization of the ising dodecahedron in a very low - field region on assumption that entropy per spin is close enough to the ground - state degeneracy at zero field @xmath81 . from this perspective
, the ising dodecahedron could be regarded together with the ising octahedron as the most prominent frustrated ising spin cluster that would allow efficient cooling through the adiabatic demagnetization .
the present work deals with the magnetization process and magnetocaloric properties of the regular ising polyhedra ( tetrahedron , octahedron , cube , icosahedron and dodecahedron ) , which are rigorously calculated by means of a relatively simple graph - theoretical approach .
apart from the exact analytic calculations , the proposed procedure is suitable also for the numerical implementation as it was illustrated on the particular example of the ising dodecahedron .
in addition , the method can be rather straightforwardly adapted in order to calculate magnetic properties of the irregular ising spin clusters with a more complex geometric structure as well .
it has been demonstrated that the magnetic behaviour of antiferromagnetic ising spin clusters with the shape of regular polyhedra is basically influenced through a geometric spin frustration with exception of the ising cube , which is the only bipartite regular polyhedron . in a consequence of that ,
the magnetization curve of the ising cube involves just one trivial plateau at zero magnetization due to a simple antiferromagnetic ground state in contrast to the magnetization curves of all other regular ising polyhedra . among these ,
the ising tetrahedron and icosahedron still exhibit a zero magnetization plateau , which is successively followed by the other intermediate one - half plateau in the former case and respectively , the intermediate one - third and one - half plateaux in the latter case . on the other hand ,
there is a lack of zero magnetization plateau in the magnetization curves of the ising octahedron and dodecahedron , which display only a single intermediate plateau at one - third and one - fifth of the saturation magnetization , respectively .
the intermediate magnetization plateaux emerging in a low - temperature magnetization process owing to a geometric spin frustration reveal highly degenerate ground states , the microscopic nature of which was clarified along with their respective degeneracies .
we have found convincing evidence that the higher number of magnetization plateaux does not necessarily imply a quantum nature of higher - order magnetization plateaux contrary to conventional expectations @xcite , but it may be just a simple consequence of geometric spin frustration .
it has been also evidenced that the magnetization jumps accompanying a transition between different magnetization plateaux manifest themselves in a giant magnetocaloric effect during the adiabatic demagnetization . from this point of view , the ising octahedron and dodecahedron have turned out as the most prominent geometrically frustrated spin clusters that would allow an efficient cooling down to lowest achievable temperatures .
100 r. l. carlin , magnetochemistry , springer , berlin , 1986 .
o. kahn , molecular magnetism , wiley , new york , 1993 .
d. gatteschi , r. sessoli , j. villain , molecular nanomagnets , oxford university press , oxford , 2006 .
a. furrer , o. waldmann , rev .
( 2013 ) 367 .
leuenberg , d. loss , nature 410 ( 2001 ) 789 .
f. troiani , a. ghirri , m. affronte , s. carretta , p. santini , g. amoretti , s. piligkos , g. timco , r.e.p .
winpenny , phys .
94 ( 2005 ) 207208 .
friedman , m.p .
sarachik , annu .
rev . condens .
matter phys . 1 ( 2010 ) 109 .
parkinson , j. timonen , j. phys . : condens .
matter 12 ( 2000 ) 8669 .
j. schnack , h .- j .
schmidt , j. richter , j. schulenburg , eur .
j. b 24 ( 2001 ) 475 .
r. schmidt , j. richter , j. schnack , j. magn .
295 ( 2005 ) 164 .
schrder , h. nojiri , j. schnack , p. hage , m. luban , p. kgerler , phys .
94 ( 2005 ) 017205 .
schrder , h .- j .
schmidt , j. schnack , m. luban , phys .
94 ( 2005 ) 207203 .
konstantinidis , phys .
b 72 ( 2005 )
064453 . j. schnack , j. low .
142 ( 2006 ) 279 .
konstantinidis , phys .
b 76 ( 2007 ) 104434 . i. rousochatzakis , a.m. luchli , f. mila , phys . rev .
b 77 ( 2008 ) 094420 .
j. schnack , r. schnalle , polyhedron 28 ( 2009 ) 1620 .
konstantinidis , phys .
b 80 ( 2009 ) 134427 . s. sahoo , a. hucht , s. sil , p. entel , aip conf . proc . 1461 ( 2011 ) 199 .
j. ummethum , j. schnack , a.m. luchli , j. magn .
327 ( 2013 ) 103 .
y. shapira , v. bindilatti , j. appl .
92 ( 2002 ) 4155 .
efremov , r.a .
klemm , phys .
b 74 ( 2006 ) 064408 . r.a .
klemm , d.v .
efremov , phys .
b 77 ( 2008 ) 184410 .
a. hucht , s. sahoo , s. sil , p. entel , phys .
b 84 ( 2011 ) 104438 .
r. liebmann , statistical mechanics of periodic frustrated ising systems , lecture notes in physics , vol .
251 , springer , berlin , 1986 .
diep , frustrated spin systems , world scientific , singapore , 2004 .
i. syozi , rev .
kobe univ .
mercantile marine ( japan ) ( 1955 ) 2125 .
fisher , phys .
( 1959 ) 969 .
vogel , j. valds , j. cartes , nanostructured materials 3 ( 1993 ) 441 .
valds , j. cartes , e.e .
vogel , rev .
46 ( 2000 ) 348 .
w. lebrecht , e.e .
vogel , j. cartes , j.f .
valds , physica a 342 ( 2004 ) 90 .
e. viitala , j. merikoski , j. timonen , m. manninen , z. phys .
d 40 ( 1997 ) 173 .
e. viitala , j. merikoski , m. manninen , j. timonen , phys .
b 55 ( 1997 ) 11541 .
m. ukovi , a. bobk , phys .
lett . a 378 ( 2014 ) 1773 .
m. ukovi , j. magn .
374 ( 2015 ) 22 .
s. yamamoto , t. sakai , phys .
b 62 ( 2000 ) 3795 .
t. sakai , s. yamamoto , j. phys .
: condens .
matter 12 ( 2000 ) 9787 . | magnetization process and adiabatic demagnetization of the antiferromagnetic ising spin clusters with the shape of regular polyhedra ( platonic solids ) are exactly examined within the framework of a simple graph - theoretical approach .
while the ising cube as the only unfrustrated ( bipartite ) spin cluster shows just one trivial plateau at zero magnetization , the other regular ising polyhedra ( tetrahedron , octahedron , icosahedron and dodecahedron ) additionally display either one or two intermediate plateaux at fractional values of the saturation magnetization .
the nature of highly degenerate ground states emergent at intermediate plateaux owing to a geometric frustration is clarified .
it is evidenced that the regular ising polyhedra exhibit a giant magnetocaloric effect in a vicinity of magnetization jumps , whereas the ising octahedron and dodecahedron belong to the most prominent geometrically frustrated spin clusters that enable an efficient low - temperature refrigeration by the process of adiabatic demagnetization .
ising spin clusters , regular polyhedra , magnetization plateaux , giant magnetocaloric effect , exact results 05.50.+q , 75.10.hk , 75.10.jm , 75.30.sg , 75.40.cx , 75.60.ej |
some time ago monopoles in einstein - yang - mills - higgs(eymh ) model , for @xmath2 gauge group with higgs field in adjoint representation , were studied as a generalization of the t hooft - ployakov monopole to see the effect of gravity on it . in particular , it was found that solutions exist up to some critical value of a dimensionless parameter @xmath1 , characterising the strength of the gravitational interaction , above which there is no regular solution .
the existance of these solutions were also proved analytically for the case of infinite higgs mass .
also , non abelian magnetically charged black hole solutions were shown to exist in this model for both finite as well as infinite value of the coupling constant for higgs field .
the abelian black holes exists for @xmath3 and non abelian black holes exist in a limited region of the @xmath4 plane .
recently born - infeld theory has received wide publicity , especially in the context of string theory .
bogomolnyi - prasad - sommerfield ( bps ) saturated solutions were obtained in abelian higgs model as well as in @xmath5 sigma model in @xmath6 dimensions in presence of born - infeld term .
different models for domain wall , vortex and monopole solutions , containing the born - infeld lagrangian were constructed in such a way that the self - dual equations are identical with the corresponding yang - mills - higgs model .
recently non self - dual monopole solutions were found numerically in non abelian born - infeld - higgs theory . in this paper
we consider the einstein - born - infeld - higgs(ebih ) model and study the monopole and black hole solutions .
the solutions are qualitatively similar to those of eymh model .
the black hole configurations have nonzero non abelian field strength and hence they are called non abelian black holes . in sec .
ii we consider the model and find the equations of motion for static spherically symmetric fields . in sec iii
we find the asymptotic behaviours and discuss the numerical results .
finally we conclude the results in sec .
we consider the following einstein - born - infeld - higgs action for @xmath2 fields with the higgs field in the adjoint representation s = d^4x with l_g & = & , + l_h & = & - d_^a d^^a -(^a^a - v^2 ) ^2 and the non abelian born - infeld lagrangian , l_bi = ^2 str ( 1 - ) where d_^a = _
^a + e ^abc a_^b^c , f _ = f_^a t^a = ( _ a_^a - _ a_^a + e ^abca_^ba_^c)t^a and
the symmetric trace is defined as str(t_1,t_2 ...
,t_n ) = tr(t_i_1t_i_2 ... t_i_n ) . here
the sum is over all permutations on the product of the @xmath7 generators @xmath8 .
here we are interested in purely magnetic configurations , hence we have @xmath9 . expanding the square root in powers of @xmath10 and keeping up to order
@xmath10 we have the born - infeld lagrangian l_bi = -f_^a f^a + + o ( ) .
for static spherical symmetric solutions , the metric can be parametrized as ds^2 = -e ^2(r)dt^2 + e ^2(r)dr^2 + r^2(r)(d^2 + ^2d^2 ) and we consider the following ansatz for the gauge and scalar fields a_t^a(r ) = 0 = a_r^a , a_^a = e_^a , a_^a = -e_^a , and ^a = e_r^a v h(r ) . putting the above ansatz in eq.1 , defining @xmath11 and rescaling @xmath12 and @xmath13 we get the following expression for the lagrangian dr e^+ , where v_1 = ( w)^2 + r^2(h)^2 - ( w)^2 , v_2 = and v_3 = + w^2h^2 + ( h^2 - 1)^2 - . here
the prime denotes differentiation with respect to @xmath14 .
the dimensionless parameter @xmath1 can be expressed as the mass ratio = with the gauge field mass @xmath15 and the planck mass @xmath16 .
note that the higgs mass @xmath17 . in the limit of @xmath18 the above action reduces to that of the einstein - yang - mills - higgs model .
for the case of @xmath19 we must have @xmath20 which corresponds to the flat space born - infeld - higgs theory .
we now consider the gauge @xmath21 , corresponding to the schwarzschild - like coordinates and rename @xmath22 .
we define @xmath23 and @xmath24 . varying the matter field lagrangian with respect to the metric we find the energy - momentum tensor . integrating the @xmath25 component of the energy - momentum
we get the mass of the monopole equal to @xmath26 where m = ^2 _ 0^ dr ( nv_1 - n^2v_2 + v_3 ) following t hooft the electromagnetic @xmath27 field strength @xmath28 can be defined as _ = - ^abc^ad_^bd_^c .
then using the ansatz(3 ) the magnetic field b^i = ^ijkf_jk is equal to @xmath29 with a total flux @xmath30 and unit magnetic charge .
the @xmath25 and @xmath31 components of einstein s equations are & & ( 1 - ( rn) ) = ^2 ( n v_1 - n^2 v_2 + v_3 ) + & & = ( v_1 - 2nv_2 ) .
the equations for the matter fields are & & ( anv_4) = a w ( ( w^2 - 1 ) + 2 h^2 - - ( w^2 - 1 ) ) + & & ( anr^2h) = a h ( 2w^2 + g^2r^2(h^2 - 1 ) ) with v_4 = 2w - ( w^2 - 1)^2 - ( w)^3 it is easy to see that @xmath32 can be elliminated from the matter field equations using eq.(12 ) . hence we have to solve three differential equations eqs .
( 11),(13 ) and ( 14 ) for the three fields @xmath33 and @xmath34 .
for finite @xmath35 , demanding the solutions to be regular and the monopole mass to be finite gives the following behaviour near the origin & & h = a r + o(r^3 ) , + & & w = 1 - b r^2 + o(r^4 ) , + & & n = 1 - c r^2 + o(r^4 ) , where @xmath36 and @xmath37 are free parameters and @xmath38 is given by c = ^2 ( a^2 + 4b^2 + - ) .
in general , with these initial conditions @xmath39 can be zero at some finite @xmath40 where the solutions become singular . in order to avoid this singularity we have to adjust the parameters @xmath36 and @xmath37 suitably . for @xmath41
we require the solutions to be asymptotically flat .
hence we impose n = 1 - then for finite mass configuration we have the following expressions for the gauge and the higgs fields & & w = c r^-m e^-r(1 + o ( ) ) + & & h = \ { ll 1 - b r^-gm - 1 e^-gr , & for 0 < g + 1 - r^-2m-2 e^-2r , & for g = 0 and g > . .
note that the fields have similar kind of asymptotic behaviour in the eymh model .
we have solved the equations of motion numerically with the boundary conditions given by eqs.(16 - 21 ) . for @xmath42 , @xmath43 and @xmath18
they corresponds to the exact prasad - sommerfield solution .
for nonzero @xmath44 and finite @xmath45 the qualitative behaviour of the solutions are similar to the corresponding solutions of eymh model . for large @xmath40
these solutions converges to their asymptotic values given as in eqs.(19 - 21 ) . for a fixed value of @xmath35 and @xmath45 we solved the equations increasing the value of @xmath1 . for small value of @xmath1 the solutions
are very close to flat space solution . as @xmath1
is increased the minimum of the metric function @xmath39 was found to be decreasing .
the solutions cease to exist for @xmath1 greater then certain critical value @xmath46 . for @xmath43 and @xmath47
we find @xmath48 .
the profile for the fields for different values of @xmath1 with @xmath49 and @xmath50 are given in figs.1,2 and 3 .
the profile for the fields for @xmath51 and @xmath52 are given in fig .
we find numerically the mass @xmath53 of the monopole for @xmath54 and @xmath50 .
apart from the regular monopoles , magnetically charged black holes can also exist in this model .
black hole arises when the field @xmath55 vanishes for some finite @xmath56 . demanding the solutions to be regular near horizon @xmath57 we find the following behaviour of the fields & & n(r_h + ) = n_h+ o(^2 ) , + & & h(r_h + ) = h_h + h_h+ o(^2 ) , + & & w(r_h + ) = w_h + w_h+ o(^2 ) with & & n_h = + & & h_h = \
{ 2w_h^2 + g^2 r_h^2 ( h_h^2 - 1 ) } + & & w_h = . here
@xmath58 and @xmath59 are arbitrary . for @xmath41
the behaviour of the fields is same as the regular monopole solution as given by eqs.(19 - 21 ) .
the black hole has unit magnetic charge with nontrivial gauge field strength .
we found numerical solutions to the non abelian black hole for different @xmath57 .
for a fixed value of @xmath60 we find the solutions for @xmath61 adjusting the parameters @xmath62 and @xmath63 .
for @xmath60 close to zero the solutions approach the regular monopole solutions . the profile for the fields are given in fig.5 .
we found the mass of the black hole equals to be @xmath64 for @xmath65 and @xmath50 .
in this paper we have investigated the effect of gravity on the born - infeld - higgs monopole .
we found that solutions exist only up to some critical value @xmath46 of the parameter @xmath1 . in the limit @xmath18 these solutions reduces to those of eymh monopoles .
we also found numerically magnetically charged non abelian black hole solutions in this model
. it would be interesting to prove analytically the existence of these solutions for finite value of the parameters .
recently dyons and dyonic black holes were found in eymh model numerically and the existence of critical value for @xmath1 was also proved analytically .
it may be possible to generalize these solutions to find dyons and dyonic black holes in ebih model .
we hope to report on this issue in future .
i am indebted to avinash khare for many helpful discussions as well as for a careful manuscript reading . 50 p. breitenlohner , p. forgace and d. maison , nucl . phys .
b 383 ( 1992 ) 357 .
m. e. ortiz , phys .
d 45 ( 1992 ) r2586 .
p. breitenlohner , p. forgace and d. maison , nucl .
b 442 ( 1995 ) 126 .
g. t hooft , nucl .
b 79 ( 1974 ) 276 ; a. m. polyakov , jetp lett . 20
( 1974 ) 194 .
k. lee , v. p. nair and e. j. weinberg , phys .
d 45 ( 1992 ) 2751 p. c. aichelburg and p. bizon phys . rev .
d 48 ( 1993 ) 607 .
m. born , proc .
london a 143 ( 1934 ) 410 .
m. born and m. infeld , proc .
london a 144 ( 1934 ) 425 .
e. s. fradkin and a. a. tseytlin , phys .
b 163 ( 1985 ) 123 ; a. a. tseytlin , nucl . phys .
b 276 , ( 1986 ) 391 ; r. g. leigh , mod . phys . lett .
a 4 ( 1989 ) 2073 ; a 4 , ( 1989 ) 2767 .
p. k. tripathy , phys .
d 59 ( 1999 ) 085004 ; s. gonorazky , c. nunez , f. a. schaposnik and g. silva , nucl . phys .
b 531 ( 1998 ) 2881 . a. nakamura and k. shiraishi , int
phys . a 6 ( 1991 ) 2635 ; hadronic j. 14 ( 1991 ) ; g. gibbons , nucl .
b 514 ( 1998 ) 603 .
n. grandi , e. f. moreno and f. a. schaposnik , hep - th/9901073 .
p. bizon , phys .
64 ( 1990 ) 2844 .
s. gonorazky , f. a. schaposnik and g. silva , phys .
b 449 ( 1999 ) 187 ; d. brecher , phys . lett .
b 442 ( 1998 ) 117 ; a. a. tseytlin , nucl . phys .
b 501 ( 1997 ) 41 .
p. g. bergmann , m. cahen and a. b. komar , j. math .
( 1965 ) 1 . m. k. prasad and c. m. sommerfield , phys
35 ( 1975 ) 760 .
y. brihaye , b. hartmann and j. kunz , phys .
b 441 ( 1998 ) 77 .
y. brihaye , b. hartmann , j. kunz and n. tell , hep - th/9904065 . | we find static spherically symmetric monopoles in einstein - born - infeld - higgs model in @xmath0 dimensions .
the solutions exist only when a parameter @xmath1 ( related to the strength of gravitational interaction ) does not exceed certain critical value .
we also discuss magnetically charged non abelian black holes in this model .
we analyse these solutions numerically . 9.0 in 6.0 in |
the 1.68-ev photoluminescence ( pl ) center was reported many decades ago in diamond@xcite and where later it was assumed @xcite that silicon impurities were involved in this center .
this was confirmed by pl measurements at cryogenic temperature on si - doped chemical vapor deposition ( cvd ) polycrystalline diamond samples where the fine structure of the 1.68-ev pl center could be detected.@xcite a 12-line fine structure is observed close to 1.68 ev , and this can be divided into three similar groups each containing four components .
the relative strengths of the optical absorption for the three groups of lines are found to be the same as the ratio of the abundancies of the natural isotopes of silicon , @xmath0si , @xmath1si , and @xmath2si@xcite .
the 4-line fine structure for an individual si - isotope is assigned to doublet levels both in the ground and excited states which split by about 48 and 242 ghz , respectively.@xcite it has been assumed that this small splitting might be explained by dynamic jahn - teller effect@xcite and/or by spin - orbit effect.@xcite recently , single photon - emission from 1.68-ev pl center has been demonstrated @xcite .
its zero - phonon - line ( zpl ) with 5 nm width even at room temperature and the near infrared emission makes this pl center very attractive candidate for quantum optics@xcite and biomarker@xcite applications .
spin - polarized local density approximation@xcite ( lda ) within density functional theory ( dft ) calculations in a very small nanodiamond model ( with @xmath370 carbon atoms ) concluded that the negatively charged split - vacancy form ( see fig .
[ fig : struct ] ) of siv defect ( siv@xmath4 ) is responsible for the 1.68-ev pl center.@xcite they exclude the neutral siv defect ( siv@xmath5 ) as a good candidate as its ground state is orbitally singlet.@xcite this model was later disputed based on a semiempirical restricted open - shell hartree - fock cyclic cluster model calculation where they claimed that tunneling of si - atom along the symmetry axis may occur for siv@xmath5 defect that can explain the doublet line in the ground state.@xcite the fingerprint of siv@xmath5 was found by electron paramagnetic resonance ( epr ) studies.@xcite the kul1 center with @xmath6 high spin ground state and d@xmath7 symmetry@xcite was recently associated with siv@xmath8 defect where 216-atom lda supercell calculations produces relatively good agreement with the measured @xmath9c and @xmath0si hyperfine couplings ( see tab .
[ tab : hf ] ) .
very recently , thorough epr and pl studies have been carried out to correlate the kul1 epr center with an 1.31-ev pl center , and its relation to the 1.68-ev pl center.@xcite the final conclusion was that siv@xmath8 has a zpl at 1.31-ev whereas siv@xmath4 has a zpl at 1.68-ev .
photo - conductivity measurements@xcite and photo - ionization measurements@xcite indicate that the adiabatic ( thermal ) charge transition level of @xmath10 level of siv defect is at @xmath11 + 1.5 ev , where vbm is the valence band edge .
while the recent measurements@xcite are very plausible still no _ ab initio _ theory was able to conclusively support the assignment of 1.68-ev center with the negatively charged siv defect .
lda or any semilocal generalized gradient approximation ( gga ) functionals suffer from the band gap error@xcite which inhibits to directly compare the calculated and experimental zpl energies .
recent advances in dft functionals made possible to accurately calculate zpl energies and charge transition level of defects in diamond and other semiconductors.@xcite in this paper , we apply this theory to study the charge transition levels and the zpl energies of siv defect in diamond .
these calculations yield the position and nature of defect states in host diamond and are able to reveal the nature of the shelving state in 1.68-ev pl center .
our paper is organized as follows . in section [
sec : method ] we describe briefly the _ ab initio _ method that we applied to study the electronic structure and excitations of the siv defect . in section [ sec : results ] we describe the structure and the basic defect level scheme of siv defect by group theory . here , we combine the results from _ ab initio _ calculations with group theory considerations in order to identify the order of important defect states and the charge state relevant for the most important 1.68-ev pl center .
we discuss the results then we conclude and summarize the results in section [ sec : summary ] .
( color online ) the split - vacancy structure of the siv defect and the calculated spin density in neutral charge state .
the isosurface of spin density is 0.05 .
the defect has @xmath12 symmetry with the symmetry axis of [ 111 ] direction .
the lattice sites of the missing carbon atoms are depicted by the smallest ( pink ) balls . ]
the present calculations have been carried out in the framework of the generalized kohn - sham theory,@xcite by using the screened hybrid functional hse06 of heyd , ernzerhof and scuseria with the original parameters ( 0.2 @xmath13 for screening and 25% mixing).@xcite hse06 in diamond happens to be nearly free of the electron self - interaction error , and is capable of providing defect - levels and defect - related electronic transitions within @xmath30.1 ev to experiment.@xcite we have used the vienna ab initio simulation package vasp 5.3.2 with the projector augmented wave@xcite ( paw ) method ( applying projectors originally supplied to the 5.2 version).@xcite to avoid size effects as much as possible , a 512-atom supercell was used in the @xmath14-approximation for defect studies .
parameters for the supercell calculations were established first by using the gga exchange of perdew , burke and ernzerhof ( pbe)@xcite in bulk calculations on the primitive cell with a 8@xmath158@xmath158 monkhorst - pack ( mp ) set for brillouin - zone sampling.@xcite ( increasing the mp set to 12@xmath1512@xmath1512 has changed the total energy by @xmath160.002 ev . ) constant volume relaxations using a cutoff of 370(740 ) ev in the plane - wave expansion for the wave function ( charge density ) resulted in an equilibrium lattice parameter of @xmath17= 3.570 . increasing the cutoff to 420 ( 840 )
ev has changed the lattice constant by only 0.003 .
therefore , considering the demands of the supercell calculations , the lower cut - off was selected . the hse06 calculation with the 8@xmath158@xmath158 mp set and 370(740 )
ev cutoff resulted in the lattice constant @xmath18=3.545 , indirect band gap of @xmath19=5.34 ev , in good agreement with the experimental values of @xmath20=3.567 and @xmath19=5.48 ev ( see , e.g. , ref . ) . due to the different choice of the basis ,
the hse06 values presented here differ somewhat from those in refs . ,
but tests on the negatively charge nitrogen - vacancy center ( nv ) have shown that the higher cutoff would cause only negligible difference in the equilibrium geometry of that defect too .
defects in the supercell were allowed to relax in constant volume till the forces were below 0.01 ev / .
we calculated the hyperfine tensors of @xmath9c and @xmath0si isotopes within paw formalism@xcite as implemented in vasp 5.3.2 package . here , we applied a larger 500(1000 ) ev cutoff for the plane wave ( charge density ) expansion .
the hyperfine tensor ( @xmath21 ) between the electron spin density @xmath22 of an electron spin @xmath23 , and the non - zero nuclear spin @xmath24 of nucleus j may be written as @xmath25 \label{eq : hyperfine}\ ] ] where the first term within the square brackets is the so - called ( isotropic ) fermi - contact term , and @xmath26 represents the ( anisotropic ) magnetic dipole - dipole contribution to the hyperfine tensor .
@xmath27 is the nuclear bohr - magneton of nucleus @xmath28 and @xmath29 the electron bohr - magneton .
the fermi - contact term is proportional to the magnitude of the electron spin density at the center of the nucleus which is equal to one third of the trace of the hyperfine tensor , @xmath30 .
the fermi contact term arises from the spin density of unpaired electrons with @xmath31-character , and can be quite sizable . the spin density built up from unpaired electrons of @xmath32-character yields the dipole - dipole hyperfine coupling .
the fraction of fermi - contact and dipole - dipole term implicitly provides information about the character of the wave function of the unpaired electron as well as the corresponding nuclei ( via @xmath33 ) . according to our recent study@xcite the contribution of the spin polarization of core electrons to the fermi - contact hyperfine interaction@xcite is significant for @xmath9c isotopes ,
thus , we include this term in the calculation of hyperfine tensors .
the excitation energies were calculated within constrained dft method that was successfully applied to nv center in diamond.@xcite in this method one can calculate the relaxation energy of the nuclei due to optical excitation .
the formation energy of the defect with defect charge state @xmath34 is defined as @xmath35 by ignoring the entropy contributions , where @xmath36 is the total energy of the defect in the supercell , @xmath37 is the calculated energy of vbm in the perfect supercell , @xmath38 and @xmath39 are the chemical potentials of c and si atoms in diamond with @xmath40 number of c atoms in the supercell , @xmath41 is the chemical potential of the electron , i.e. , the fermi - level , and @xmath42 is the correction needed for charged supercells .
the thermal charge transition level between the defect charge states of @xmath34 and @xmath43 is the position of the fermi - level ( @xmath41 ) in the fundamental band gap of diamond where the formation energies are equal in these charge states .
this condition simplifies to a difference in the total energies in their respective charge states as follows , @xmath44 where @xmath41 is referenced to the calculated @xmath37 . for comparison of different defect configurations and charge states , the electrostatic potential alignment and the charge correction scheme of lany and zunger
was applied for @xmath42.@xcite recently , this scheme was found to work best for defects with medium localization.@xcite
we first study the structure and the obtained defect levels by hse06 calculation .
then , we apply group theory in order to explain the symmetry of the defect states .
we use the calculated thermal ionization energies , excitation energies and hyperfine couplings to identify the 1.31-ev and 1.68-ev pl centers .
we also discuss the results with comparing them to the experiments .
we first calculated the neutral defect siv@xmath8 by substituting the c - atom by a si - atom adjacent to a nearby vacancy .
the si automatically left the substitutional site creating a split - vacancy configuration which may be described as a si - atom placed in a divacancy where the position of the si - atom is equidistant from the two vacant sites ( see fig .
[ fig : struct ] ) .
the position of si - atom is a bond center position which is an inversion center of perfect diamond lattice . in our special coordinate frame
the two vacant sites reside along the [ 111 ] direction of the lattice which has a @xmath45 rotation axis .
the symmetry of the defect may be described as @xmath46 or @xmath12 , where @xmath47 is the inversion .
we note that nv center has @xmath48 symmetry with no inversion .
the defect has @xmath6 high - spin ground state .
this finding agrees with the lda calculations.@xcite after establishing the symmetry of the defect one can apply group theory analysis for this defect.@xcite .
one can build the defect states of this defect as an interaction between the divacancy orbitals and the si - impurity states .
the si impurity has 6 immediate neighbor c - atoms in divacancy
. the calculated distance between si- and c - atoms is about 1.97 in the neutral charge state , which is longer than the usual si - c covalent bond ( 1.88 ) .
since c - atoms are more electronegative than si - atom the charge transfer from the si - atom toward the c - atoms can be relatively large leaving positively charged si ion behind .
the divacancy has @xmath12 symmetry with six c dangling bonds .
these dangling bonds form @xmath49 , @xmath50 , @xmath51 and @xmath52 orbitals while the si - related four @xmath53 states should form @xmath49 , @xmath51 , and @xmath50 orbitals in @xmath12 crystal field ( the explicit form of these orbitals as a function of @xmath53 states can be seen in ref . ) .
please , note that the @xmath49 , @xmath50 and @xmath51 states may be combined but @xmath52 orbitals should be pure c dangling bonds state ( see fig . [
fig : leveldiag ] ) .
the bonding and anti - bonding combinations of these states form the defect states of siv defect .
( color online ) the defect - molecule diagram of the neutral siv defect in diamond .
the irreducible representation of the orbitals under @xmath12 symmetry is shown .
the orbitals of the si - atom and the six carbon dangling bonds of divacancy recombine in diamond where the conduction band ( cb ) and valence band ( vb ) are schematically depicted .
the @xmath52 orbital is very close to the valence band maximum with forming a strong resonance state at the valence band edge shown as brown ( lighter gray ) lines .
the @xmath51 level falls in the valence band that also forms a strong resonance even deeper in the valence band . for the sake of the simplicity , the position of the spin - up ( majority spin channel ) levels
are depicted in the spin - polarized density functional theory calculation . ] according to hse06 calculation the occupation of the defect states may be described as @xmath54@xmath55@xmath56@xmath57 which agrees again with previous lda calculations.@xcite here , the ten electrons are coming from the six electrons of c dangling bonds and the 4 @xmath53 electrons of the si impurity . as two electrons occupy the double degenerate @xmath52 state , the high - spin @xmath6 ground state naturally forms by following hund s rules .
it is important to determine the position of the defect levels .
interestingly , hse06 predicts that only @xmath52 appears in the band gap . in the spin - polarized calculation
the occupied @xmath52 state in the spin - up channel is at @xmath37 + 0.3 ev .
the @xmath51 state is resonant with the valence band and can be found just 0.64 ev below vbm .
the occupied @xmath49 state lies very deep in the valence band and may play no important role in the excitation or ionization processes .
our estimations indicate that the @xmath50 defect level is too deep in the vb to be excited by red excitation .
however , higher energetic lasers ( in the green and blue spectrum ) can excite this state and other states within the vb with @xmath50 symmetry .
the empty anti - bonding orbitals fall in the conduction band , and will not be considered any more . the most important @xmath51 and @xmath52 states
are depicted in fig .
[ fig : states ] . despite the @xmath51 state lies in the valence band it is still localized around the defect site .
the valence band states are strongly perturbed by the presence of the defect .
the vbm is triple degenerate in perfect lattice that splits to an upper @xmath49 level and a lower @xmath52 level in the presence of the defect .
since this @xmath58 state has the same symmetry as the low - lying @xmath52 defect state , that @xmath58 state becomes a defect resonance state .
similar phenomena occurs for the deep @xmath51 defect state as well
. the shallow @xmath58 state may play an important role in the excitation / de - excitation process of the defect .
we conclude that the neutral siv defect ( i ) has @xmath59 ground state , ( ii ) can be theoretically ionized as ( 2 + ) , ( 1 + ) , as well as ( 1- ) and ( 2- ) by emptying or filling the double degenerate @xmath52 state in the gap . in order to establish the relevant charge states
, one has to calculate the adiabatic charge transition levels of siv defect .
we found that the neutral , negatively charged and double negatively charged states can be found in diamond ( see fig .
[ fig : occlev ] ) .
the single positive charged state might be only found in very highly p - type doped diamond samples . the calculated formation energy of siv defect as a function of the fermi - level in the gap .
the crossing points represent the charge transition levels .
the chemical potential of si is taken from cubic silicon carbide in the carbon - rich limit . ] the calculated @xmath10 level at @xmath37 + 1.43 ev is very close to the level that is associated with the acceptor ionization energy of the defect from photo - conductivity measurements at @xmath3@xmath37 + 1.5 ev.@xcite interestingly , the calculated @xmath60 level at @xmath3@xmath37 + 2.14 ev is well below the midgap .
since the @xmath52 level is fully occupied at @xmath37 + 1.5 ev in @xmath61 charge state intra - level optical transition can not take place , and ultra - violet excitation ( @xmath34.0 ev ) is needed to excite or ionize the defect optically to the conduction band edge .
another important note that the calculated acceptor level of nv lies at @xmath3@xmath37 + 2.6 ev.@xcite this means that if both nv and siv defects are present in the diamond sample then most of the nv defects should be neutral in order to detect siv@xmath62 defect .
another important point that the luminescence from siv@xmath62 defect can be more stable than that of nv@xmath62 defect in nanodiamonds as a function of surface termination because the corresponding @xmath60 charge transition level lies deeper in the band gap than the @xmath10 level of nv defect .
all in all , the neutral and negatively charged siv defects are relevant for intra - level optical transitions . in the negatively charge siv defect
the ground state electron configuration is @xmath56@xmath63 .
this is principally a jahn - teller unstable system as the double degenerate @xmath52 level is partially filled ( in the spin - down channel in our calculation ) .
this has @xmath64 symmetry in @xmath12 symmetry . in hse06
geometry optimization we allowed the systems to relax to lower symmetries .
indeed , hse06 showed a @xmath65 distortion where two c - atoms have 0.03 longer distance from si - atom than the other two atoms . in this particular case
@xmath66 state was formed in @xmath65 symmetry .
however , we have to note that the dynamic coupling between vibrations and electronic states can not be taken into account in our calculation .
thus , dynamic jahn - teller system can not be directly described by our method .
for instance , static jahn - teller effect occurs for nv@xmath8 in ref . while it is known from experiments that it is a dynamic jahn - teller system .
since the distortion from @xmath12 symmetry obtained by hse06 calculation is small the defect may well have @xmath12 symmetry with dynamic jahn - teller effect .
we note that no such an electron paramagnetic resonance center with @xmath67 spin was found that could be associated with siv - defect .
this also hints that the ground state of siv@xmath4 is a dynamic jahn - teller system which prohibits the electron spin resonance signal similar to nv@xmath8 defect .
now , we discuss the possible excited states of this system .
again , the fully occupied @xmath51 state is resonant with the valence band .
still , one can promote one electron from this level to the @xmath52 level in the band gap .
the resulted @xmath68 excited state is again jahn - teller unstable .
interestingly , when the electron from the minority spin - down @xmath51 level in the valence band was promoted to the @xmath52 state in the gap then the resulted hole state pops up clearly above the valence band edge at about @xmath37 + 0.12 ev [ see fig . [
fig : sivexc](a ) ] . the @xmath69 optical transition is allowed . in hse06 calculation the excited state
has also @xmath65 distortion but the effect is again small , and can be a dynamic jahn - teller system .
the calculated zpl energy is about 1.72 ev which agrees well with that of 1.68-ev pl center .
the calculated relaxation energy due to optical excitation is about 0.03 ev which is much smaller than that of nv center ( about 0.21 ev ) , and explains the small contribution of the vibration sideband in the emission spectrum .
thus , the assignment of 1.68-ev pl center with siv@xmath62 defect is well - supported by our calculation .
we argue that the sharp zero - phonon line of siv@xmath4 as well as the relative strength of the fine - structure splittings in the ground and excited states can be well - understood by our findings . the @xmath64 ground and @xmath68 excited states have very similar electron charge densities , where mostly just the phase differs between the two states . as a result
, the ions will be subject to similar potentials in their ground and excited states leading to only small change in the geometry due to optical excitation .
this is in stark contrast to the case of nv@xmath4 defect where the electron charge density strongly redistributes upon optical excitation leading to a large stokes - shift.@xcite another observation is that @xmath64 state virtually is not localized at all on si - atom due to symmetry reasons , however , our projected density of states analysis shows that there is a small contribution from the orbitals of the si - atom in the @xmath68 excited state , allowed by symmetry .
this may explain the larger splitting in the fine - structure of the excited state than that in the ground state:@xcite the @xmath68 state has expected to have larger spin - orbit splitting due to the small contribution of the orbitals of si - atom than that in @xmath64 state where this contribution is missing , as the spin - orbit strength increases with the atomic number to the fourth power .
( color online ) schematic diagram about the electronic structure of the negatively charged siv defect in diamond . the valence band ( vb ) resonant states play a crucial role in the excitation . for the sake of the simplicity ,
the spin - polarization of the single particle levels are not shown .
( a ) ground and the bright ( optically allowed ) excited state .
( b ) ground and shelving states when the hole left on the split valence band edge states .
( c ) schematic energy level diagram based on hse06 constraint density functional theory calculations .
thick red lines indicate strong absorption / emission while the thin red line indicates a weak radiative recombination in the case of the presence of strain which breaks the inversion symmetry of the defect .
orange wavy arrows represent non - radiative recombination down to the ground state . , title="fig : " ] + ( color online ) schematic diagram about the electronic structure of the negatively charged siv defect in diamond .
the valence band ( vb ) resonant states play a crucial role in the excitation . for the sake of the simplicity ,
the spin - polarization of the single particle levels are not shown .
( a ) ground and the bright ( optically allowed ) excited state .
( b ) ground and shelving states when the hole left on the split valence band edge states .
( c ) schematic energy level diagram based on hse06 constraint density functional theory calculations .
thick red lines indicate strong absorption / emission while the thin red line indicates a weak radiative recombination in the case of the presence of strain which breaks the inversion symmetry of the defect .
orange wavy arrows represent non - radiative recombination down to the ground state .
, title="fig : " ] + furthermore , our hse06 calculations reveal that the heavily perturbed valence band edges can play important role in understanding the defect properties as those states lie closer to the @xmath52 state in the band gap than the @xmath51 resonant defect state .
therefore , we calculate the excitation from the split vbm states .
a @xmath70 and a @xmath71 excited states are built from the hole left on the @xmath58 and @xmath72 states , respectively [ c.f . , fig .
[ fig : sivexc](b ) ] .
we find that the @xmath70 excited state is @xmath30.1 ev lower in energy than the bright @xmath68 excited state where the calculated stokes - shift of this transition is again small .
the @xmath71 excited state is @xmath31.59 ev above the @xmath70 ground state .
these results can explain the nature of the shelving state in 1.68-ev pl center .
unlike nv - center the siv defect has inversion symmetry that plays a crucial role in the de - excitation process . due to the inversion symmetry the @xmath70 and @xmath71 excited states are shelving states because they are optically forbidden due to the same parity of their wave function with that of the ground state .
thus , the shelving states have the same spin state as the ground and the bright excited states .
since @xmath70 shelving state has a level very close to that of the bright @xmath73 state the non - radiative coupling between these states can be efficient .
we further note that a new and weak near - infrared ( nir ) transition at @xmath3823 nm ( 1.52 ev ) has been found associated with the negatively charged siv - defect@xcite .
the measurements implied that this transition belongs to the same charge as the 1.68-ev transition .
this nir transition was particularly found in ensemble measurements of nanodiamond samples when strain was present in the sample@xcite .
our calculations imply [ fig .
[ fig : sivexc](c ) ] that this weak radiative transition can be explained by the slightly distorted @xmath71 excited state .
any distortion of the diamond lattice will break the inversion symmetry of the siv defect , so the parity of the corresponding wave functions .
therefore , transitions between @xmath74 and @xmath75 ( originally @xmath76 and @xmath64 ) will be allowed .
the calculated transition energy ( 1.59 ev ) is very close to the detected one which further supports our assignment .
our calculations highlight the importance of the valence band edge states in understanding the optical properties of 1.68-ev center in diamond .
we note again that the @xmath51 state is resonant with the valence band .
thus , unlike the case of nv center with well - separated atomic - like states in the gap , it is probable that ionization of the defect can occur during optical excitation . in this process
, a hole is created that might ( temporarily ) leave the defect with creating an optically inactive ( @xmath77 ) charge state , particularly , in the presence of external perturbations that create an effective electric field . this ( @xmath77 ) charge state can be optically converted back to ( @xmath78 ) state only by ultra - violet excitation .
the ( @xmath77 ) charge state is a closed - shell singlet while the neutral and negatively charged siv defects have @xmath6 and @xmath67 spin states , respectively .
the siv@xmath8 defect was assigned to kul1 epr center with @xmath6 state and @xmath12 symmetry.@xcite .
our calculations supports this assignment ( see table [ tab : hf ] ) .
the spin density is localized on four c dangling bonds in the @xmath52 orbital of divacancy .
while the @xmath52 state is not localized on si impurity but the spin density from these c dangling bonds can overlap with the si atom which promotes a well measurable fermi - contact term on the @xmath1si nuclei .
the hyperfine interaction with @xmath1si is almost isotropic unlike the case of @xmath9c isotopes which show typical anisotropic signal due to @xmath53 dangling bonds .
.[tab : hf ] calculated and measured hyperfine constants ( @xmath79 ) for kul1 epr center and neutral silicon - vacancy defect in diamond .
the kul1 epr center data was taken from ref . .
previous theoretical results are taken from ref . carried out by lda 216-atom supercell calculation without taken into account the core polarization ( lda - nocp ) .
the present theoretical values are obtained by hse06 functional in 512-atom supercell where the contribution of the spin - polarization of core electrons to the fermi - contact term in @xmath9c is 26% of the total ( hse06-cp ) .
@xmath80 is the angle between the symmetry axis ( [ 111 ] direction ) and the parallel component of the hyperfine constant @xmath81 . [
cols="<,^,^,^",options="header " , ] @xmath82 ref . ,
@xmath83 ref .
the calculated and measured @xmath9c hyperfine tensor agree very well .
we note that the contribution of the spin - polarization of core electrons to the fermi - contact term is very significant which _ compensates _ the hyperfine interaction due to valence electrons .
we found this effect also for nv and other related defects.@xcite next , we discuss the optical transitions for siv@xmath8
. intra - level transition may occur between the fully occupied @xmath51 state and the empty @xmath52 state in the spin - down channel .
while the fully occupied @xmath51 state lies in the valence band a strong resonant excitation may occur from this state . in hole picture
the excited state may be described as @xmath84@xmath85 while the ground state is @xmath57 .
the electron configuration of @xmath57 results in @xmath86 triplet state and @xmath87 , @xmath88 singlet states .
the ground state is @xmath86 .
the electron configurations of @xmath84@xmath85 results in @xmath89 , @xmath90 , @xmath91 , @xmath92 , @xmath93 , and @xmath94 multiplets . among these states @xmath89 and @xmath93 states are optically allowed from the @xmath86 ground state .
polarization studies of the 1.31-ev center indicates that @xmath95 transition occurs in the pl process@xcite .
the @xmath89 state can be described as linear combination of slater - determinants that can not be treated within constraint dft . however
, the @xmath96=1 substate of @xmath93 multiplet can be described by a single slater - determinant .
the calculated excitation energy is about 1.63 ev which is significantly larger than 1.31 ev .
thus , we may conclude that @xmath97 transition is not responsible for the 1.31-ev pl line . as the only other feasible transition is @xmath95
, our result indirectly confirms the experimental finding .
we also study the role of the valence band resonant states similar to the case of the negatively charged defect .
when the electron is promoted from @xmath72 state to the @xmath52 state in the gap then @xmath98 state is created .
the calculated excitation energy of this shelving state is @xmath31.27 ev .
this coincides with the small peak found in ref . , simultaneously with the peak of 1.31-ev
this transition is optically allowed only when the inversion symmetry of the defect is broken .
this might happen due to strain in the diamond sample .
nevertheless , our calculation implies that this @xmath98 state can play an important role in the de - excitation process as a shelving or metastable state .
we note that the experiments indicate@xcite that the ground state of the neutral siv defect resides at @xmath30.2 ev above the valence band edge . according to hse06 calculations
the occupied single particle @xmath52 level lies at @xmath37 + 0.3 ev that agrees nicely with the implications from the experiments .
in addition , we found a metastable state of the defect which may explain the temperature dependence of the 1.31-ev pl spectrum@xcite .
we also propose that strain may induce a weak pl transition from this shelving state because this shelving state has the same spin as that of the ground state , and the parity selection rule may be relaxed in slightly distorted geometry of the siv defect .
we further note that a very similar conclusion can be drawn for the optical excitation of the neutral siv defect as was hinted for the negatively charged defect : a hole is created in the valence band in the excitation process that may leave defect temporarily or permanently which leads to a charge conversion from neutral to negatively charged siv defect .
we carried out _ ab initio _ supercell calculations on siv defect in diamond .
the calculations could positively confirm the assignment of 1.68-ev and 1.31-ev pl centers to the negatively charged and neutral siv defect , respectively .
the calculations reveal the high importance of the inversion symmetry of the center as well as the role of resonant valence band states in understanding the optical properties of the defect .
we show that the shelving states of 1.68-ev pl center are from valence band excitations where the lowest - energy shelving state may have nir emission to the ground state in strain diamond samples .
we show that holes are created in the excitation of the negatively charged and neutral siv defects that may lead to charge conversion of these centers .
in addition , we show that the acceptor level of siv defect lies very deep in the band gap . as a consequence , the 1.68-ev pl center can be photo - stable when it is close to the surface of hydrogenated diamond surface .
on the other hand , siv defect is double negatively charged in such diamond samples where the nv - defect is negatively charged .
thus , the fermi - level should be set lower than the midgap of diamond , in order to conserve its negative charge state .
discussions with christoph becher are highly appreciated .
support from eu fp7 project diamant ( grant no .
270197 ) is acknowledged .
j.r.m . acknowledges support from conicyt pia program act1108 .
39ifxundefined [ 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 * * , ( ) link:\doibase 10.1103/physrevb.51.16681 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.77.3041 [ * * , ( ) ] http://stacks.iop.org/0953-4075/39/i=1/a=005 [ * * , ( ) ] http://stacks.iop.org/1367-2630/13/i=2/a=025012 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.205211 [ * * , ( ) ] http://stacks.iop.org/1367-2630/15/i=4/a=043005 [ * * , ( ) ] link:\doibase 10.1039/b813515k [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1134/1.1626778 [ * * , ( ) ] link:\doibase 10.1103/physrevb.62.16587 [ * * , ( ) ] link:\doibase 10.1103/physrevb.66.195207 [ * * , ( ) ] link:\doibase 10.1103/physrevb.77.245205 [ * * , ( ) ] link:\doibase 10.1103/physrevb.84.245208 [ * * , ( ) ] link:\doibase 10.1063/1.358566 [ * * , ( ) ] link:\doibase 10.1002/pssb.201046254 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.103.186404 [ * * , ( ) ] link:\doibase 10.1103/physrevb.81.153203 [ * * , ( ) ] @noop * * ( ) link:\doibase 10.1103/physrevb.76.115109 [ * * , ( ) ] link:\doibase 10.1063/1.1564060 [ * * , ( ) ] link:\doibase 10.1063/1.2404663 [ * * , ( ) ] link:\doibase 10.1103/physrevb.50.17953 [ * * , ( ) ] link:\doibase 10.1103/physrevb.54.11169 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.77.3865 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1002/pssb.201046210 [ * * , ( ) ] link:\doibase 10.1103/physrevb.62.6158 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevb.71.115110 [ * * , ( ) ] link:\doibase 10.1103/physrevb.78.235104 [ * * , ( ) ] link:\doibase 10.1103/physrevb.86.045112 [ * * , ( ) ] link:\doibase 10.1103/physrevb.76.075204 [ * * , ( ) ] http://arxiv.org/abs/1310.3106 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1103/physrevb.79.235210 [ * * , ( ) ] link:\doibase 10.1103/physrevb.85.245207 [ * * , ( ) ] | the split silicon - vacancy defect ( siv ) in diamond is an electrically and optically active color center .
recently , it has been shown that this color center is bright and can be detected at the single defect level .
in addition , the siv defect shows a non - zero electronic spin ground state that potentially makes this defect an alternative candidate for quantum optics and metrology applications beside the well - known nitrogen - vacancy color center in diamond .
however , the electronic structure of the defect , the nature of optical excitations and other related properties are not well - understood . here
we present advanced _
ab initio
_ study on siv defect in diamond .
we determine the formation energies , charge transition levels and the nature of excitations of the defect .
our study unravel the origin of the dark or shelving state for the negatively charged siv defect associated with the 1.68-ev photoluminescence center . |
the threshold between the classical and quantum world has fascinated physicists since the discovery of quantum phenomena and realizing how different they are from our everyday experience .
one of the prominent examples of quantum behavior is the nonlocality leading to violation of bell s inequalities @xcite . for two - level systems
there is no nonlocality without quantum entanglement , but the opposite can be true @xcite .
quantum entanglement plays an important role in quantum information processing @xcite .
however , the entanglement is not the only type of non - classical correlations . as described by ollivier and zurek @xcite
the nonclassical correlations can be associated with _
quantum discord_. quantum discord ( qd ) is useful in many ways including quantum information processing or detection of quantum phase transitions , especially in the cases when the entanglement fails to grasp this phenomenon @xcite .
moreover , it was demonstrated that only classical correlations can be broadcast locally @xcite .
all of these features of quantum discord motivate the quest for developing tools for detecting and quantifying it .
nevertheless , there were only a few experimental implementations of witnesses of nonclassical correlations , or _
nonclassicality witnesses _ ( nws ) , in discrete - variable systems .
two of them were implemented in nuclear magnetic resonance systems @xcite and one using linear optics @xcite , however these witnesses were not universal . at this point
, we should stress that detecting purely classical correlations is a difficult problem since it involves solving optimization problem over a nonconvex set of classical states .
thus , the problem of detecting classical correlations is harder that detection of entanglement .
moreover , any nw should be nonlinear @xcite .
for those reasons the nws @xcite are usually non - universal . however , zhang @xcite demonstrated that finding a universal nw is possible , but the established witness is not suitable for optical implementation . a better suited qd indicator ( qdi ) which overcomes the limitations of all the previously developed witnesses is a measure introduced by girolami and adesso @xcite .
we call it an indicator instead of a witness since in contrast to a typical witness it is universal and on average its value provides a bound on qd .
let us start with introducing some basic definitions used throughout our paper .
a general two - qubit density matrix @xmath0 can be expressed in the bloch representation as @xmath1 where @xmath2 $ ] and matrix @xmath3 $ ] are given in terms of the pauli matrices , and @xmath4 $ ] ( @xmath5 $ ] ) describe bloch vector @xmath6 ( @xmath7 ) of the first ( second ) subsystem , later referred to as @xmath8 and @xmath9 . moreover , it is always possible to transform @xmath0 with local unitary operations @xcite so that @xmath10 becomes a diagonal matrix .
the state @xmath0 is not entangled ( is separable ) when it has a positive partial transpose , i.e. , is a ppt state ( see peres - horodecki criterion @xcite ) .
the lack of entanglement for a two - qubit system implies , e.g. , locality , in terms of violation of the bell - chsh inequality @xcite ( for quantitative study see @xcite ) , and thus it corresponds to classical situation where the measurement outcomes can be explained by a hidden - variable model . however , quantum entanglement is not responsible for all the nonclassical effects .
one of the recently celebrated manifestation of quantumness is _ quantum discord _ @xcite .
the qd is responsible for the difference in conditional quantum information calculated in two ways , where one of them uses the bayesian rule for calculating probabilities .
therefore , qd quantifies how much conditional quantum probabilities differ from those calculated within classical theory .
the qd vanishes if the state fulfills the strong ppt condition @xcite , i.e. , @xmath0 has to be ppt and its ppt must admit cholesky decomposition ( there are also other so - called nullity conditions for review see @xcite ) .
thus , if there is no discord , there is no entanglement .
however , the reverse does not have to be true .
there are several ways of quantifying qd .
the one for which an analytic formula is known @xcite is the so - called _ geometric quantum discord _ ( gqd ) quantifying hilbert - schmidt distance to the closest non - discordant state .
the expression for the gqd reads @xmath11 where @xmath12 ( for @xmath13 ) stand for eigenvalues of matrix @xmath14 or @xmath15 , where @xmath16 denotes transposition .
the largest @xmath12 is denoted as @xmath17 .
note that @xmath18 is asymmetric .
thus , if @xmath19 the state is called classical quantum or if @xmath20 the state is quantum - classical .
naturally , there have been attempts of finding an analytic formula for the symmetric gqd , which answers the question about the closest classical - classical state , however this is still an open problem @xcite . if @xmath21 the state is classical - classical since it does not exhibit quantum correlations responsible for discord between conditional quantum information calculated in the two above - mentioned ways . in the following sections we show how to experimentally identify states of zero @xmath18 and
describe how to perform the experiment within the framework of linear - optics by measuring @xmath22 introduced in in ref .
the qdi provides a tight and faithful lower bound ( @xmath23 ) for gqd and reads @xmath24\leq d_i , \label{eq : witnessa}\ ] ] where @xmath25 for @xmath26 are moments of the matrix @xmath27 ( @xmath13 ) from eq .
( [ eq : g_discord ] ) , where @xmath28 denotes @xmath29th eigenvalue of @xmath27 .
note that @xmath22 and @xmath27 are asymmetric , thus @xmath22 can not exclusively detect classical - classical states .
one of the possible symmetric qdis is @xmath30 .
moreover , since the symmetric geometric discord vanishes @xmath31 only if @xmath21 @xcite , for checking if the state exhibits classical - classical correlations @xmath32 should be used . finally , let us note that there are other qdis than @xmath22 that are functions of the moments @xmath33 for @xmath26 providing a faithful bound on geometric quantum discord , e.g. , @xmath34 however , as demonstrated in fig .
[ fig:1 ] , @xmath22 provides a better estimation of @xmath18 .
( color online ) monte carlo simulation outcomes for @xmath35 random two - qubit density matrices .
geometric quantum discord @xmath18 versus normalized quantum discord indicators @xmath22 and @xmath36 . , width=302 ] ( color online ) diagram for measuring @xmath37 ( left ) and @xmath38 ( right ) for a two - qubit source emitting states @xmath0 with period @xmath39 . for both qubits from a pair
the delay varies depending on the pair number in the sequence . in the case of measuring @xmath40 the qubits in arm @xmath8 ( @xmath9 ) are delayed by @xmath41 or @xmath39 ( @xmath39 or @xmath42 ) , whereas for @xmath43 the qubits in arm @xmath8 ( @xmath9 ) are delayed by @xmath44 , @xmath39 , or @xmath42 ( @xmath39 or @xmath42 ) .
the difficulty of implementing the delay in a real experiment depends on the value of @xmath39 and on technical aspects depending on the physical properties of the qubits . for a very small @xmath39
it can be difficult to provide fast enough switching of the delay , whereas for a very large @xmath39 coherent storage of qubits could be challenging.,title="fig : " ] ( color online ) diagram for measuring @xmath37 ( left ) and @xmath38 ( right ) for a two - qubit source emitting states @xmath0 with period @xmath39 .
for both qubits from a pair the delay varies depending on the pair number in the sequence . in the case of measuring @xmath40 the qubits in arm @xmath8 ( @xmath9 ) are delayed by @xmath41 or @xmath39 ( @xmath39 or @xmath42 ) , whereas for @xmath43 the qubits in arm @xmath8 ( @xmath9 ) are delayed by @xmath44 , @xmath39 , or @xmath42 ( @xmath39 or @xmath42 ) .
the difficulty of implementing the delay in a real experiment depends on the value of @xmath39 and on technical aspects depending on the physical properties of the qubits . for a very small @xmath39
it can be difficult to provide fast enough switching of the delay , whereas for a very large @xmath39 coherent storage of qubits could be challenging.,title="fig : " ]
up to this point we did not mention how to measure @xmath45 .
the proposed measurement procedure is similar to the proposal for direct measurement of quantum discord introduced by jin in @xcite and it resembles the so - called entanglement swapping @xcite . in our case we need only the first two moments of the matrix @xmath27 , and we need fewer copies of @xmath0 .
the first moment is given as @xmath46 , where we fixed without loss of generality @xmath47 , and requires two copies of @xmath0 to be measured .
this is because @xmath37 can be also expressed @xcite as @xmath48,\label{eq : m1}\end{aligned}\ ] ] where @xmath49 both @xmath50 and @xmath51 are given in terms of singlet projections @xmath52 and @xmath53 with @xmath13 and @xmath54 .
however , for measuring @xmath55 we need four copies @xmath0 , since @xmath56.\label{eq : m2}\end{aligned}\ ] ] the operators @xmath57 and @xmath58 are two - qubit operators acting on the respective subsystems . having four copies of @xmath0 each moment can be estimated in a single coincidence measurement . however , if six copies are available , the nw from eq .
( [ eq : witnessa ] ) could be evaluated in a single measurement .
we can however , use temporal separation between the copies ( as shown in figs .
[ fig:2 ] and [ fig:3 ] ) instead of spatial separation to investigate the quantum correlations of a two - qubit source of a constant frequency of @xmath59 .
the qdi from eq .
( [ eq : witnessa ] ) can be measured in a linear - optical system where the qubit @xmath60-basis states @xmath61 and @xmath62 are represented by single - photon horizontal @xmath63 and vertical @xmath64 polarization states .
our proposal for the experimental setup consists of standard optical elements , i.e. , 50:50 asymmetric beam splitters ( bs ) and photon detectors .
since the qubits are polarization encoded it may seem surprising that we do not use polarization - dependent components as , e.g. , polarizing bss , however the polarization dependence that we take advantage of is induced in the bosonic commutation relations that are fulfilled by the bs transformed photons . as shown in fig .
[ fig:2 ] we need a two - photon source to conduct the experiment given that we are able to implement the correct sequence of delays for the corresponding photons .
this task could can be very demanding and therefore we propose an alternative experiential procedure using a four - photon source as shown in fig .
[ fig:3 ] .
high quality multiphoton sources of are difficult to obtain ( for a review see ref .
@xcite ) but implementing them is easier than providing the complex reliable and efficient system of delays as described in fig .
[ fig:2 ] .
( color online ) diagram of measurement of @xmath55 using four photons ( two two - qubit states @xmath0 ) emitted at intervals @xmath39 . from left : @xmath65 iterations of the measurement procedure employing deterministic delay ; a single iteration of the procedure in case of the probabilistic delay . in the latter case
the time required for obtaining a single measurement outcome is twice as long as in the other case since in order to obtain the correct result one needs to ensure that photons left the delay lines .
thus , the deterministic scheme is @xmath66 times more efficient than the probabilistic one , where @xmath67 is the success probability of the delay .
however , the deterministic scheme requires switching the duration of the delay with frequency @xmath59 equal to the repetition rate of the source .
the measurement diagram for @xmath37 for a four - photon source is trivial since it does not require delays .
, title="fig : " ] ( color online ) diagram of measurement of @xmath55 using four photons ( two two - qubit states @xmath0 ) emitted at intervals @xmath39 . from left : @xmath65 iterations of the measurement procedure employing deterministic delay ; a single iteration of the procedure in case of the probabilistic delay . in the latter case
the time required for obtaining a single measurement outcome is twice as long as in the other case since in order to obtain the correct result one needs to ensure that photons left the delay lines .
thus , the deterministic scheme is @xmath66 times more efficient than the probabilistic one , where @xmath67 is the success probability of the delay .
however , the deterministic scheme requires switching the duration of the delay with frequency @xmath59 equal to the repetition rate of the source .
the measurement diagram for @xmath37 for a four - photon source is trivial since it does not require delays .
, title="fig : " ] ( color online ) setup for linear - optical measurement of @xmath68 or @xmath69 observables , i.e. , the @xmath70 box .
the setup consists of 50:50 asymmetric beam splitters ( bss ) and conventional detectors ( @xmath71 ) . to measure @xmath68 ( @xmath69 ) a value of @xmath72
is assigned to coincidence @xmath73 at @xmath74 and @xmath75 as it implies singlet projection [ see eq .
( [ eq : c23 ] ) ] and value of 1 ( 2 ) if the coincidence @xmath76 implying @xmath77 operation [ see eq .
( [ eq : c14 ] ) ] is registered by @xmath78 and @xmath79 .
the other combinations of detector clicks are irrelevant , thus we do not assign values to other detection events .
the device works with probability @xmath80 because with this probability both photons reach unaltered the central bs or exit to detectors @xmath78 and @xmath79.,width=226 ] the operators constituting moments @xmath37 and @xmath55 are @xmath57 and @xmath58 .
each of the operators is a linear combination of the projections @xmath81 and identity operations @xmath82 .
both of these fundamental operations can be implemented using linear optics ( see fig .
[ fig:4 ] ) with polarization encoded qubits .
operation @xmath82 corresponds to detecting unaltered photons .
the singlet projections @xmath81 is implemented by detecting a pair of photons in separate spatial modes , say 2 and 3 , after they interacted on a bs assuming that both photons impinge on the bs from two separate modes , e.g. , @xmath29 and @xmath83 , as well . the asymmetric 50:50 bs transmits and
reflects a photon with probability @xmath80 , but the photon gets a @xmath84 phase shift only if it reflects from one side side of the bs ( marked by a gray layer in fig . [ fig:4 ] ) . note that the bs transformation is unitary and is its own inverse , i.e. , @xmath85 .
first let us show that the two - mode identity operator @xmath77 can be implemented as a coincidence @xmath76 at detectors @xmath78 and @xmath79 shown in fig .
[ fig:4 ] .
this coincidence can be expressed as @xmath86 where @xmath87 is quantum efficiency of the detectors @xmath88 and @xmath89 .
we neglect the dark counts since depending on the wavelength and the detectors used they can be very rare , moreover the dark counts are uncorrelated and thus their simultaneous appearance is highly improbable .
once we trace back the photons detected by @xmath76 to the input modes @xmath90 of the setup from fig . [ fig:4 ] by performing the bs transformation we obtain @xmath91,\end{aligned}\ ] ] where @xmath92 . since no photons arrive form the vacuum modes @xmath93 and @xmath94 the @xmath76 operator effectively reads as @xmath95 or in polarization basis where @xmath96 it reads @xmath97 thus , by performing coincidence detection after the two bss we measure an operator proportional to @xmath77 , where the constant coefficient @xmath98 is known in advance
the analogous reasoning as in the case of the two - mode identity operator @xmath99 can be applied to show that coincidence @xmath73 performs the @xmath100 projection .
let us start with the operator describing the coincidence at detectors @xmath74 and @xmath75 @xmath101 which can be rewritten in the bell basis @xmath102 the bell states are transformed by the bs as follows : @xmath103 reads @xmath104 note that all the bs transformed bell states except the singlet state have two - photon components which are impossible to appear because there can be only one photon per mode @xmath105 and @xmath106 .
thus , we can simplify the operator @xmath107 describing the bs transformation followed by coincidence detection to @xmath108 which in terms of the input modes reads @xmath109 where the terms describing the detection of photons emerging from vacuum modes @xmath93 and @xmath94 were not taken into account .
the operators @xmath68 and @xmath69 are linear combinations of @xmath110 and @xmath111 of the following form @xmath112 the factor of @xmath113 compensates for the fact that @xmath110 or @xmath111 are both measured probabilistically with a success rate of @xmath92 , but the total success rate @xmath114 of the @xmath70 setup is @xmath115 .
the probabilistic nature of the setup reduces uniformly the number of all the coincidence counts by a factor of @xmath98 .
thus , we can express the average values of @xmath116 and @xmath117 as @xmath118 where @xmath119 is the number of registered coincidences @xmath76 and @xmath120 is the number of registered coincidences @xmath73 .
estimating the averages of products of the @xmath50 and @xmath51 operators is a bit more involved since it requires using more than one @xmath70 block shown in fig .
[ fig:4 ] .
our experimental proposal based on the probabilistic scheme from fig . [ fig:3 ] shown in fig . [ fig:5 ] uses two such blocks .
let us assume that we perform @xmath65 iterations of the experiment for @xmath123 .
for each iteration the @xmath70 box provides us with one of three values @xmath124 for measuring @xmath68 and @xmath125 for measuring @xmath69 .
the there values ( @xmath126 or @xmath127 ) correspond to not observing coincidence @xmath76 or @xmath73 , observing coincidence @xmath76 , and observing coincidence @xmath73 , correspondingly .
thus , for measuring @xmath128 the @xmath83th measurement outcome is a product of two numbers and reads @xmath129 , whereas for @xmath130 it is a product of four numbers , i.e. , @xmath131 .
we can express the expectation values as @xmath132 where the number of events where @xmath133 or @xmath134 was measured is @xmath135 and @xmath136 , correspondingly , where @xmath137 is the kronecker s @xmath137 and @xmath67 is the success probability of the delay .
( color online ) optical setup for measuring @xmath122 using the probabilistic scheme from fig .
[ fig:3 ] with @xmath138 . when the delay lines are removed , the setup measures @xmath121 . at times
@xmath139 and @xmath140 two photon pairs @xmath0 enter the setup .
when @xmath141 no new photons appear from sources @xmath142 and @xmath143 and only coincidences in the @xmath50 part of the setup are witnessed .
the photons are delayed before entering the @xmath50 part .
one of the photons is always delayed by @xmath39 and the second one randomly by @xmath41 or @xmath42 , however the photon with probability @xmath144 can leak out of the system . in the successful cases @xmath145 and
@xmath146 enter the setup at @xmath139 and @xmath147 is measured .
at the same time photons in the left part are delayed , the first from the left by @xmath39 and the second by @xmath41 ( with probability @xmath148 ) , so at @xmath139 there is no photon detected in @xmath50 .
photon pairs @xmath149 and @xmath150 enter the setup at @xmath140 and @xmath151 is measured without delay . in the left part , the first photon is delayed by @xmath39 , but the second one is not affected ( with probability @xmath148 ) . at the same time
@xmath152 is measured .
since there are no photons added at @xmath153 only @xmath154 is measured ( for a photon provided at @xmath139 and delayed by @xmath41 and a photon provided at @xmath140 delayed by @xmath39 ) .
the success rate of the delay procedure is @xmath155 .
the whole sequence is repeated until a good estimate of @xmath122 is obtained , e.g. , we reach @xmath156 successful iterations .
no photons enter the setup at @xmath157 to not affect the next measurement iteration.,width=302 ] note that for the case of measuring @xmath37 and @xmath55 we lose @xmath158 and @xmath159 photons , correspondingly . in spite of losing many photons measuring @xmath22
is expected to be faster than full 2-qubit _ quantum tomography _
( qt ) which requires rotating a number of polarization plates providing 16 measurement configurations , where each rotation takes a few seconds . during this time for small @xmath39 the measurement setup from fig
. [ fig:5 ] would accumulate enough data to estimate @xmath22 .
this assessment , however , strongly depends on the efficiency of the detectors @xmath87 ( see fig .
[ fig:6 ] ) .
time @xmath160 in units of @xmath39 ( the inverse of time between generation of the two two - photon states ) needed for performing @xmath161 successful measurements of @xmath121 and @xmath122 versus the detection efficiency @xmath87 .
the success rate of generating two pairs of photons from a single pump pulse was assumed to be @xmath162 .
the solid curve corresponds to the probabilistic delay scheme whereas the dashed one corresponds to the deterministic delay scheme .
the repetition rate of the pump can be set to @xmath163 ( @xmath164 ) . te repetition rate is limited by the dead time of the detectors ( typically about @xmath165 ) .
measuring both moments with @xmath161 iterations in a time of order of @xmath166 would be possible for perfect detectors and deterministic delay scheme . in a realistic case of @xmath167 and probabilistic
delay the whole measurement would take less than @xmath168 . note , however , that @xmath169 estimates for @xmath121 and @xmath122 would be available in a 10 times shorter time.,width=264 ] the main difficulty in implementing the setup outlined in fig . [ fig:5 ] is working with two two - photon sources @xmath170 and @xmath171 .
the photon pairs should exhibit quantum correlations , thus should be produced , e.g. , in spontaneous parametric down - conversion of light of a certain degree of depolarization . since the pairs are produced by a random process , the coincidence rate in our experimental proposal can be very low .
however , as we demonstrate below , it is not always the case .
there are of course further challenges .
depending on the approach of producing the input states , one may face the problem of indistinguishability of the photon pairs , which causes the results to differ from the expected one , but methods of circumventing this effect can be found , e.g. , in @xcite .
moreover , the theoretical prediction is reached if all the pairs perfectly overlap on the corresponding bss .
this condition is easy to be satisfied for one pair of photons , however the amount of work increases with the number of photons . to achieve this
all the optical elements would require active stabilization . the success rate of the measurement of @xmath122 from fig .
[ fig:5 ] is @xmath172 since it requires applying probabilistic delay for measuring @xmath50 . when all the delay lines in fig .
[ fig:5 ] are removed , the setup and can be used for direct measurement of @xmath121 with success rate of @xmath173 .
the success rates should also incorporate the probabilities of generating photon pairs at @xmath174 . in two - crystal type - i geometry ( so - called kwiat source @xcite ) , the probability of generating an entangled photon pair from a single pump pulse is about @xmath175 .
thus , creating two pairs independently occurs with probability @xmath162 and the effective success probabilities read @xmath176 and @xmath177 for @xmath121 and @xmath122 , correspondingly . by taking into account the pauses in providing two pairs of photons at @xmath178 ( we only allow every second two pulses from the pulse train to enter the setup by using a pulse picker ) the efficiency of the setup for measuring @xmath122 drops by a factor of @xmath80 .
thus , the useful coincidence count rate for @xmath122 would be limited to @xmath179 of the brightness of the source .
the repetition rate of the source ( @xmath59 ) is limited by the dead time of the detectors ( about 50ns ) to about 20mhz .
thus , for realistic detector efficiency of @xmath180 ( perkin - elmer single photon counting modules operating at wavelength of 700 nm ) , the number of successful @xmath122 measurements per second would reach @xmath181 and @xmath182 for @xmath121 .
these numbers would be further reduced by a few percent by unavoidable imperfections of the setup . nevertheless , the final expected number of detection events for @xmath122 is of order of @xmath183 coincidences per second which would allow us to measure @xmath122 in less than three minutes . measuring @xmath121 would be much faster because of the simpler structure of coincidences , no pulse picking , and lack of probabilistic delay . the deterministic setup for measuring @xmath122 outlined in fig .
[ fig:3 ] would also not suffer from such a low coincidence rate as the setup shown in fig .
[ fig:5 ] .
however , the later alternative is more feasible and can be successfully implemented , e.g. , in our laboratory . for the purpose of our experiment
we need two two - photon sources @xmath171 and @xmath170 in fig .
[ fig:5 ] producing photon pairs in the same state @xmath0 .
however , the sources can produce slightly different states @xmath184 and @xmath185 for @xmath171 and @xmath170 , correspondingly . in this case
we assume that our state @xmath0 is given as @xmath186 .
if the states @xmath184 and @xmath185 are not identical the fidelity @xcite of these two states @xmath187\right\rbrace^2\ ] ] is less than @xmath188 and the measured moments read @xmath189,\label{eq : dm1}\end{aligned}\ ] ] and @xmath190.\label{eq : dm2}\end{aligned}\ ] ] using the moments ( [ eq : dm1 ] ) and ( [ eq : dm2 ] ) for calculating @xmath191 we obtain @xmath192 which varies with fidelity @xmath193 of the sources @xmath171 and @xmath170 . the value of @xmath194 can be successfully used to approximate @xmath195 , where @xmath186 . for reasonably large fidelity values
@xmath196 this approximation is expected to be very accurate since for the @xmath35 monte carlo generated states @xmath184 and @xmath185 shown in fig .
[ fig:7 ] the largest module of the difference @xmath197 is smaller than 0.05 , i.e. , @xmath198 . obtaining fidelity above @xmath199 should not be difficult , thus , the setup can be considered robust to imperfections of the sources . the measured value of @xmath194 for the two sources @xmath171 and @xmath170 producing pairs of photons in states @xmath184 and @xmath185 versus the precise @xmath191 value for @xmath200 .
the blue data points are monte carlo results for @xmath35 randomly generated @xmath185 and @xmath184 states satisfying @xmath196 .
the blue line corresponds indicates @xmath201 .
, width=264 ]
we demonstrated that @xmath45 of girolami and adesso can be efficiently measured in linear - optical system utilizing two sources of two - photon states . due to its properties
the qdi which does not require any prior knowledge about the investigated state @xmath45 is more powerful than the other recently reported nws @xcite .
the qdi and the nw introduced by zhang in @xcite both require four copies of the input state , but the latter one is not appropriate for optical implementation .
our approach is less complex than full qt , since it only requires two measurements , if four copies of @xmath0 are available [ either spatially ( see fig . [ fig:2 ] ) or time separated ( see fig .
[ fig:3 ] ) ] , instead of 16 for the qt @xcite .
furthermore , the proposed measurement setup is expected to be faster than the direct qt , since it would provide the outcome within a minute .
let us note that if the one had access to six copies of @xmath0 , the qid could be estimated in a single measurement in contrast to the simpler experiment described in @xcite , where the nw was not universal and its estimation involved measuring three quantities .
our approach can be extended ( by using larger number of copies ) to directly measure gqd as discussed in ref .
moreover , our results can help to establish discord related nws based on matrices of moments for optical fields ( studied previously in the context of entanglement @xcite and other nonclassical phenomena @xcite ) .
finally , let us note that the discussed setup can measure moments @xmath121 and @xmath122 which can constitute other qdis or nws ( see ref .
@xcite ) , thus it can be used for investigating various quantum properties of light in a way that was previously available only for the nmr platform .
the authors thank a. miranowicz for fruitful discussions .
the authors gratefully acknowledge the support by the operational program research and development for innovations
european regional development fund ( project no .
cz.1.05/2.1.00/03.0058 ) and the operational program education for competitiveness - european social fund ( projects no .
cz.1.07/2.3.00/20.0017 , no .
cz.1.07/2.3,00/20,0058 , no .
cz.1.07/2.3.00/30.0004 , and no .
cz.1.07/2.3.00/30.0041 ) of the ministry of education , youth and sports of the czech republic .
the authors also acknowledge support by the internal grant agency of palacky university in olomouc ( project no .
prf_2012_003 ) .
k.b . was supported by grant no .
dec-2011/03/b / st2/01903 of the polish national science centre . | the threshold between classical and nonclassical two - qubit states is drawn at the place when these states can no longer be described by classical correlations , i.e. , quantum discord or entanglement appear .
however , to check if the correlations are classical ( in terms of quantum discord and entanglement ) it is sufficient to witness the lack of quantum discord because its zero value implies the lack of entanglement .
we explain how the indicator of quantum discord introduced by girolami and adesso [ phys .
rev .
lett . * 108 * , 150403 ( 2012 ) ] can be practically measured in linear - optical systems using standard beam splitters and photon detectors .
we study the efficiency of the setup assuming both ideal and real components and show that the efficiency of the proposed implementation is better than the full two - photon quantum tomography .
thus , we demonstrate that a class of experiments previously available on nmr platform can be implemented in optical systems . |
quantum walks ( qws ) @xcite are especially interesting because of their importance in quantum algorithms research @xcite and also because they represent an intriguing quantum version of the ubiquitous classical phenomenon of random walks ( rws ) .
originally conceived as a single walker whose left or right step on a line is entangled with the outcome of flipping a single two - sided quantum coin , the qw has been extended to more general cases of higher - dimensional walks @xcite , multiple walkers and/or multiple coins @xcite .
these generalizations enable exploration of qws in complex settings , which could connect with real - world phenomena such as transport in photosynthetic protein complexes @xcite .
additionally the inclusion of more walkers and/or coins can efficiently enlarge the hilbert space being explored because a linear increase in degrees of freedom exponentially enlarges the hilbert space . here
we explore the complex case of two walkers each carrying and ( quantumly ) flipping separate coins but with the freedom to swap partially the coins between flips .
by partial swapping we mean that the two walkers can effect the unitary operation swap@xmath0 on their coins : for @xmath1 , this corresponds to no swapping whatsoever so the two walkers evolve completely independently , and the other extreme is @xmath2 corresponding to completely swapping the coins after every step . for @xmath3 a consequence of full or partial
swapping is to cause entanglement between the walkers position and coin degrees of freedom . classically swapping coins does not change the walker dynamics because each coin flip is random , but quantumly the effect of swapping is important because of the unitary dynamics .
partial or full swapping of coins is interesting as this simple procedure turns independent qws into entangled qws .
as multi - walker qws could occur in nature , our work suggests new phenomena to explore in qw behavior .
although we focus on the two - walker case , the multiple - walker case can be obtained by a natural extension of our foundational work on this area . full and partial
swapping of coins between independent qwers is also appealing practically as an implementation .
each independent walker s can correspond to the phase of a cavity field , and each walker s two - sided coin is an atom in the cavity with a superposition of ground and excited electronic energy states coupled to the cavity via a dispersive nonlinear phase shift @xcite .
the atoms can interact via a dipole - dipole interaction @xcite between ( or during ) steps , and the strength of this dipole - dipole interaction determines the value of @xmath4 .
two - walker qws are compared to two - walker rws by examining the joint position distributions of the two walkers after tracing out the coin states .
classically , for any value of @xmath4 , no correlation is expected and the diffusion of position is marked by its spread @xmath5 ( standard deviation of the position distribution ) increasing as @xmath6 for @xmath7 the elapsed time ( which can be expressed as a continuous or a discrete quantity and is the discrete number of steps in our study ) .
quantum walks can be distinguished by ballistic spreading , where the @xmath5 grows linearly with @xmath7 @xcite or by anderson - type localization where the walker s spread becomes constant at large @xmath7 @xcite .
this localization effect is due to the walker effectively becoming trapped due to irregularities of the spacing of the lattice traversed by the walker .
thus , there are three domains of spreading in the asymptotic large @xmath7 limit : ballistic ( @xmath8 ) , diffusive ( @xmath9 ) and localized ( @xmath10constant ) . here
we are interested not only in how the walker s position spreads but also in the correlations between two walkers positions after tracing out the coins . to study the correlations , we go beyond evaluating joint position distributions for different values of @xmath7 to studying two - time correlation functions .
such correlation functions prove to be quite useful for studying two - walker dynamics with partial coin swapping .
for a single - walker qw on a line , the hilbert space is @xmath11 with the walker hilbert space @xmath12 spanned by the orthogonal lattice position vectors @xmath13 and @xmath14 the two - dimensional coin space spanned by two orthogonal vectors which we denote @xmath15 .
each step by the walker is effected by two subsequent unitary operators : the coin - flip operator @xmath16 for @xmath17 the hadamard matrix and the conditional - translation operator @xmath18 the resultant step operator is @xmath19 with @xmath20 the identity operator on @xmath12 .
the choice of initial state @xmath21 is important in studies of qws because the interference features sensitively depend on the choice of state .
this sensitivity is persistent because the dynamics are unitary hence do not die out . on the other hand the general properties of qws do not depend on the choice of initial state so the choice of initial state is not crucial provided that the focus is on such characterization .
as we are interested in general properties , the initial state is not important so we choose the initial product state with the walker at the origin of the two - dimensional lattice and hold separate coins in an equal superposition of the @xmath22 and @xmath23 states : @xmath24 the differences between qws and rws are due to interference effects ( or entanglement ) between coin and walkers after several applications of @xmath25 ( with the number of applications @xmath7 being the discrete time variable ) .
we perform a measurement at some point in order to know the outcome of the walk .
the measurement of the walker s position corresponds to the projection - valued measure @xmath26 resulting in the walker s state ` collapsing ' to position @xmath27 on the lattice .
after @xmath7 steps , the final state of the system is @xmath28 .
the reduced state of the walker is obtained by tracing out the coin .
the probability @xmath29 that the walker will be found at the position @xmath27 is @xmath30 which is obtained by tracing out the coin of the walker - coin state and then measuring the walker s position .
we can characterize @xmath29 by the moments of this position distribution @xmath31 .
the mean @xmath32 and variance @xmath33 can be used as the measure of qws and show the signature of qws compared to rws . for a rw , the position variance @xmath34 , which is characteristic of diffusive motion ,
whereas , for a qw , a quadratic enhancement is achieved : the position variance @xmath35 @xcite .
the extension to two independent quantum walkers is straightforward .
the new hilbert space is @xmath36 , and the step operator is @xmath37 .
the walkers evolve independently of each other . to entangle the walkers and their respective coins together , we change the step operator from @xmath37 to @xmath38 for @xmath39 expressed in the basis that naturally extends from the single - walker case via the tensor product of hilbert spaces .
the full initial state of the two - walker system is @xmath40 ( each walker is initially localized at the origin of their respective lines ) .
after @xmath7 steps , the final state of the two - walker+coins system is @xmath41 with @xmath42 .
we now follow the evolution by performing a fourier transform of the evolution operator to the `` momentum '' @xmath43 space .
the eigenvectors @xmath44 of @xmath45 and @xmath46 in eq .
( [ eq : s ] ) have the eigenrelations @xmath47 for @xmath43 a continuous real quantity .
the inverse transformation is @xmath48 each walker is initialized at the origin of the line so each walker s initial state is @xmath49 in the @xmath50 and @xmath51 bases for the two wakers , where @xmath52 is a continuous index for the momentum of the second walker , the evolution operator becomes @xmath53 with @xmath54 the coin state and @xmath55 where @xmath56 the @xmath43 and @xmath52 subscripts are of course the continuous walker - momentum eigenvalues .
the general density operator for the initial state of the system in the @xmath43 basis can be expressed as @xmath57 the final state after @xmath7 steps is @xmath58 in terms of the superoperator @xmath59 , @xmath60 the walkers positions are on a two - dimensional integer lattice labeled @xmath61 with the initial position localized at @xmath62 .
measurement of the first walker s position corresponds to the projection - valued measure @xmath63 on the two - walker reduced state ( after tracing out the coins ) , and the second walker s position measurement is @xmath64 .
joint measurement of the two walkers positions corresponds to the projection - valued measure @xmath65 acting on the reduced two - walker state .
the joint position distribution @xmath66 of finding the first walker at @xmath27 and the second at @xmath67 is @xmath68 at time @xmath7 where @xmath69 is the state of the two walkers obtained by tracing out the coins .
thus the position distribution of each walker at position @xmath27 ( the case for @xmath67 is similar ) is obtained by tracing out the states of the two coins and the other walker @xmath70 the variance of the position distribution is especially important . for standard qws with single walkers , the variance of the position distribution has been shown to evolve according to @xmath71 , whereas @xmath8 for rws .
this quadratic speedup of spreading in a unitary evolution is a hallmark of the qw on a line . for our case , namely qws with two walkers switching their coin partially for each step
, the variance of each walker distribution can still be used to qualify the quantum behavior of this kind walks .
if we trace out the second walker , the reduced density matrix is @xmath72 the sum can be exactly carried out in terms of derivatives of the @xmath73 function : @xmath74 inserting this result back into the expression of the reduced density matrix yields @xmath75 the probability for the first walker to reach a point @xmath27 at time @xmath7 is @xmath76 thus we can calculate the moments of this distribution : @xmath77 for the first moment , we obtain @xmath78 where @xmath79
. we can carry out a similar integration by parts to obtain the second moment : @xmath80 + \sum_{l=1}^t\sum_{l'=1}^{l-1 } \text{tr}\left[z_1\mathcal{l}^{l - l'}_{kj}\left(\left(\mathcal{l}_{kj}^{l'}{\left|{\phi_0}\right\rangle}{\left\langle{\phi_0}\right|}\right)z_1\right)\right]\right\}.\ ] ] for @xmath1 , the two - walker qw turns to be the standard hadamard walk for each walker .
the first and second moments of the distribution behave as same as those for the standard hadamard walk .
that is , @xmath81 is linear dependent on @xmath7 , while @xmath82 is quadratic dependent on @xmath7 . for @xmath83 , @xmath84 is still unitary . for short time , the moments of the position distribution behave similarly to those for standard unitary walk . in the long - time limit
we can solve the moments of distribution analytically .
suppose @xmath85 is a linear transformation and all the eigenvalues satisfy @xmath86 .
we drop the terms that will be zero at large @xmath7 from the expressions for the moments . as @xmath85 is linear ,
we can represent it as a matrix acting on the space of @xmath87 operators .
we choose the representation to be @xmath88 the action of @xmath85 on @xmath89 is given by the matrix @xmath90 with none - zero matrix elements @xmath91,\nonumber\\ & ( m_{kj})_{1,4}=(m_{kj})_{4,1}=(m_{kj})_{13,16}=(m_{kj})_{16,13}= \frac{1}{8}e^{-i(k+j)}\left\{2+\left[1+(-1)^\tau\right]e^{2ij}-\left[1+(-1)^\tau+2e^{2ij}\right]e^{2ik}\right\},\nonumber\\ & ( m_{kj})_{1,13}=(m_{kj})_{4,16}=(m_{kj})_{13,1}=(m_{kj})_{16,4}= \frac{1}{8}e^{-i(k+j)}\left\{2-\left[1+(-1)^\tau\right]e^{2ij}+\left[1+(-1)^\tau-2e^{2ij}\right]e^{2ik}\right\},\nonumber\\ & ( m_{kj})_{1,16}=(m_{kj})_{4,13}=(m_{kj})_{13,4}=(m_{kj})_{16,1}= \frac{1}{4}\left[(1+e^{i\pi \tau})\cos(k - j)+2\cos(k+j)\right],\nonumber\\ & ( m_{kj})_{6,6}=i(m_{kj})_{6,7}=-i(m_{kj})_{7,6}=(m_{kj})_{7,7}=i(m_{kj})_{10,6}=-(m_{kj})_{10,7}=(m_{kj})_{11,6}=i(m_{kj})_{11,7}\nonumber\\&= -\frac{1}{4}\left[-1+(-1)^\tau\right]e^{i(j - k)},\nonumber\\ & ( m_{kj})_{6,10}=i(m_{kj})_{6,11}=i(m_{kj})_{7,10}=-(m_{kj})_{7,11}=-i(m_{kj})_{10,10}=(m_{kj})_{10,11}=(m_{kj})_{11,10}=i(m_{kj})_{11,11}\nonumber\\&= i\frac{1}{4}\left[-1+(-1)^\tau\right]e^{i(j - k)}.\end{aligned}\ ] ] in the integrand , the initial density matrix for the coins is multiplied @xmath92 times by @xmath85 then multiplied on the left by @xmath93 and finally the trace is taken .
as @xmath94 , @xmath95 is the same as multiplying the vector @xmath96 @xmath92 times by @xmath97 and then keeping only the @xmath98 , @xmath99,@xmath100 and @xmath101 components of the result .
this gives the new expression of the first moment as @xmath102 which is independent on @xmath7 in the long time limit since @xmath103 with all the eigenvalues of @xmath97 satisfying @xmath104 .
we then obtain @xmath105\right\}t+\text{oscillatory terms}.\ ] ] for the second moment of the distribution , the superoperator @xmath85 is unitary and preserves the identity , @xmath106 .
we will separate the traceless part @xmath107 of the coins initial state from the overall state @xcite , @xmath108 then insert this result into the expression for the second moment of the distribution to obtain @xmath109 + 2r_1\text{tr}\left[z_1\sum_{l=1}^t\sum_{l'=1}^{l-1}\mathcal{l}_{kj}^{(l - l')}z_1\right]\nonumber\\ & + \sum_{l=1}^t\sum_{l'=1}^{l-1}\text{tr}\left[z_1\mathcal{l}^{l - l'}_{kj}\left(z_1(\mathcal{l}^l_{kj}\chi_0)+(\mathcal{l}^l_{kj}\chi_0)z_1\right)\right]\big\}\approx c_2t^2+\text{const.}+\text{oscillatory terms},\end{aligned}\ ] ] where @xmath110.\ ] ] in brief the first and second moments of the position distribution are linearly and quadratically dependent on @xmath7 for @xmath1 . for @xmath83 , in the long - time limit , a ballistic behavior for each walker is obtained .
thus in the long - time limit the variance of the position distribution for one of the walkers is quadratically dependent on @xmath7 . in our case
the walker cycles through a finite sequence of coins ( two ) for each step , we still get ballistic transport . if the coin is chosen randomly from step to step , i.e. choosing coin from a infinite sequence for each step , a diffusive spreading will be obtained @xcite .
in this section we analyze the two - walker state after reducing the density matrix by tracing out the two coins . in subsec . [
subsec : joint ] we calculate the joint probability distribution for two walkers to show the correlation introduced by the swap operation on coins . in subsec .
[ subsec : mutual ] we characterization classical and total correlation using the mutual information ( mi ) and the quantum mutual information ( qmi ) , respectively , between the two coin - sharing walkers . in subsec . [ subsec : corr ] , we employ measurement - induced disturbance ( mid ) as the measure of choice for quantifying quantum correlations between the two walkers .
the joint probability distribution @xmath66 is calculated for finding walker @xmath111 in position @xmath27 and walker @xmath112 in position @xmath67 for time @xmath7 and shown in fig .
[ figure1 ] . steps with various swap - operator powers : ( a ) @xmath1 , ( b ) @xmath113 , ( c ) @xmath114 , and ( d ) @xmath2.,width=321 ] for the power of the swap operation @xmath1 , which means the two walkers are uncorrelated , @xmath66 is simply the product of the two independent one - walker distributions , @xmath115 , where @xmath116 is the probability distribution for finding walker @xmath111 in position @xmath27 after @xmath7 and similarly for @xmath117 and walker @xmath112 . the correlation functions @xmath118 are shown in fig .
[ figure1 ] , the products of two independent single - walker distributions is evident in fig .
[ figure1 ] , which shows the @xmath1 case .
however , for @xmath3 , the walkers are not independent of each other due to the walkers ( fractionally ) swapping their coins after each step .
evidently the swap operation on the two coins correlates and perhaps entangles the two coins . for classical rws
, the correlation between the coins do not affect the walkers behavior and the position distribution of each walker keeps the same as that of a standard rw with one walker .
compared to rws , the interference effect and entanglement between each walker and his coin make the two - walker qw behave differently .
the correlation between the coins is transferred to the walkers thereby inducing correlation between the two walkers . however , the interference effect and entanglement between each walker and his coin is partially destroyed by the swap operation . in fig .
[ figure1 ] , the joint position distributions of two walkers after the @xmath119 step with various powers @xmath4 of the swap operation @xmath4 displays key concepts . for @xmath1 ,
the two walkers are completely independent . for increasing @xmath4
, a small peak appears in the middle of the distribution , which means there is correlation between the two walkers .
for @xmath2 , the peak in the middle gets large and the distribution still shows the quantum behavior of two walkers . thus the correlation between two walkers increases with increasing power @xmath4 of the swap operation . in probability theory and information theory ,
the mutual information of two random variables quantifies mutual dependence of the two random variables .
here we use the mutual information as a measure of classical correlation between two quantum walkers fractionally swapping their coins . formally the mutual information of the positions of the two walkers can be defined as @xmath120\text{d}x\text{d}y .
\label{eq : mi}\end{aligned}\ ] ] as there is no classical correlation , the mutual information of two independent walkers is zero .
the swap operation between two coins connects the two walkers and introduces correlations .
we use the mutual information as a measure of classical correlations between two walkers . in fig .
[ fig : mi ] the mutual information of the two walkers positions is depicted for various powers of the swap operation @xmath4 . with various choices of the swap operator power @xmath4.,width=321 ] in the case of @xmath1 , the two walkers , each holding an independent hadamard coin , walk on a separate line independently of the other s dynamics so the mutual information of the two walkers positions is necessarily zero . for a classical rw ,
if two walkers exchange their coin after every step , any correlation between each walker and his coin is severed .
however , for the qw , correlations between the two walkers are not completely destroyed . for @xmath121
the two coins are maximally entangled by the @xmath122 operation , thereby maximizing the correlation between the two walkers .
in addition to classical correlations , the swap operation on two coins introduces quantum correlations between the two walkers as well .
now we consider the total correlation between two quantum walkers on a line under the influence of the swap operation between the coins .
we use the quantum mutual information as a measure of the total correlation including classical and quantum correlations @xcite between the walkers . given a bipartite state @xmath123 and the reduced density matrices denoted by @xmath124 and @xmath125 , the quantum mutual information , @xmath126 = s\left[\rho_1(t)\right]+s\left[\rho_2(t)\right]-s\left[\rho_\text{w}(t)\right ] , \label{eq : mutual}\ ] ] is a reasonable measure for total correlation between systems @xmath111 and @xmath112 , where @xmath127 denotes von neumann entropy .
the quantum mutual information can be interpreted analogously to the classical case , namely @xmath126=s\left[\rho_\text{w}(t)\|\rho_{1}(t)\otimes\rho_{2}(t)\right ] , \label{eq : qmi}\ ] ] where @xmath128 denotes quantum relative entropy . intuitively , the quantum mutual information reports the shared quantum information between walkers @xmath111 and @xmath112 .
quantum mutual information is a good measure for quantifying the reduction of the uncertainty of one variable through knowing the other variable .
for example , if walkers @xmath111 and @xmath112 are independent , then knowing walker @xmath111 s position does not give any information about @xmath112 s position and vice versa so their quantum mutual information is zero . otherwise , if walkers @xmath111 and @xmath112 are correlated , the quantum mutual information is positive , and the bound on this correlation is approximately @xmath129 with @xmath130 the dimension of the walker state .
the total correlation between two walkers increases with the powers @xmath4 of the swap operations as does the classical correlation .
figure [ fig : qmi ] presents plots of the quantum mutual information for an evolving two - walker qw system with various powers of the swap operation . with various powers of the swap operator @xmath131$].,width=321 ] for fixed value of @xmath4 , the quantum mutual information increases monotonically as the number of the steps increases . for @xmath1 ,
the quantum mutual informations stays at zero , which means the total correlations are cut off between the two walkers .
the quantum mutual informations for @xmath3 are positive and a maximum for @xmath121 .
we can see the maximal correlation ( both total correlation and classical correlation ) between walkers where we expect to see the coins entangled maximally . quantum discord @xcite is a popular measure for characterizing purely quantum correlations within a bipartite state , but evaluating quantum discord requires extremizing over local measurement strategies .
such an evaluation can be intractable so measurement - induced disturbance was proposed as an operational measure that avoids such an onerous optimization .
measurement - induced disturbance relies on diagonalizing the reduced density operators , which is tractable for small systems . for standard qws
, measurement - induced disturbance acts as a loose upper bound on quantum discord and tends to reflect well behavioral trends of the quantum discord @xcite .
measurement - induced disturbance and quantum discord have been compared for noisy linear and cyclic qws @xcite .
whereas measurement - induced disturbance has an operational definition , operational meaning for quantum discord is less straightforward @xcite .
on the other hand measurement - induced disturbance tends to overestimate non - classicality because it does not optimize over local measurements . despite this over - estimation , measurement - induced disturbance gives a loose upper bound on quantum discord and reflects well quantum correlations for qws @xcite .
nevertheless , care is need in interpretations of quantum correlations from measurement - induced disturbance as regimes exist where lack of optimization leads to overestimates of non - classicality .
given the reduced two - walker bipartite state @xmath123 ( tracing over coins ) , let the reduced density matrices for the two walkers be diagonalized to @xmath132 for @xmath133 , where @xmath134 is a complete projection - valued measure ( i.e. using von neumann measurements ) for each @xmath135 .
summing over joint projections on the two - walker state yields the diagonalized state @xmath136 the leftmost @xmath137 is an operator on the density matrix that diagonalizes it in the spectral basis corresponding to @xmath138 projective measurements .
the operator @xmath137 can also be described as a ` local measurement strategy ' .
correlations between the two reduced - walker states @xmath124 and @xmath125 are regarded as classical if there is a unique local measurement strategy @xmath137 leaving @xmath139 unaltered from the original two - walker state @xmath123 @xcite .
we ascertain whether the reduced two - walker state is ` quantum ' by determining whether a local measurement strategy exists that leaves the two - walker state unchanged .
the degree of quantumness is given by the measurement - induced disturbance @xcite @xmath140=i\left[\rho_\text{w}(t)\right]-i\left[\pi\rho_\text{w}(t)\right],\ ] ] for @xmath141 the quantum mutual information . by construction
@xmath139 is classical .
hence @xmath142 $ ] quantifies the classical correlations in @xmath143 , which must equal @xmath144 ( [ eq : mi ] ) .
thus , the measurement - induced disturbance is the difference between the quantum and classical mutual information , which quantify total and classical correlations .
accordingly , eq . ( [ eq : mid ] )
is interpreted as the difference between the total and classical correlations , which are represented by the quantum mutual information and the mutual information . with various powers of the swap operation
@xmath131$].,width=321 ] figure [ fig : mid ] presents plots of measurement - induced disturbance of two walkers in a two - walker qw system . from the relationship between total , classical and quantum correlations discussed in the previous paragraph
, we expect a tight quantitative relationship between figs .
[ fig : mi ] , [ fig : qmi ] and [ fig : mid ] .
we see that the total correlation represented by quantum mutual information in fig .
[ fig : qmi ] dwarfs the classical correlation represented by mutual information in fig .
[ fig : mi ] .
hence figs .
[ fig : qmi ] and [ fig : mid ] are quite similar .
pure quantumness , represented by the measurement - induced disturbance depicted in fig .
[ fig : mid ] , is monotonically increasing with @xmath4 , and pure quantum correlations thus exist between the two walkers .
this pure quantum correlation is not due to direct walker - walker interactions but are rather due to the walkers fractionally swapping their coins after each step , and the pure quantum correlations survive the tracing out of these coins .
we analyze the dynamics and entanglement of two quantum walkers who fractionally swap ( i.e. perform a swap@xmath0 unitary operation with @xmath145 ) coins .
we use mutual information , quantum mutual information and measurement - induced disturbance as measures to isolate classical vs quantum correlations .
quantum discord would be a valuable measure to use but is not tractable in our case .
however , measurement - induced disturbance suffices to show that pure quantum correlations are induced by having the walkers fractionally swap coins and then tracing out the degrees of freedom for the two coins .
in fact both classical and quantum correlations co - exist between the two walkers for @xmath3 .
quantum correlations are strongest for @xmath121 where the @xmath122 has maximal entangling power and is applied after each step .
fractional swapping is a quintessentially quantum process .
classically the walkers can swap their coins or not , or they could swap their coins sometimes , either deterministically or probabilistically . in our two - walker fractional - coin - swapping scenario , a fractional coin swap is effected after every step identically . classically one would expect that coin swapping would not affect the dynamics anyway because the coin state is unimportant : the flip chooses a random outcome .
our classical understanding of coin swapping thus gives little intuition about the quantum case .
thus , our analysis is quite valuable in that we characterize this two - walker fractional - coin - swapping case carefully and devise appropriate , tractable , meaningful classical and quantum correlation measures to study entanglement for this system .
we thank c. di franco , a. ahlbrecht , a. h. werner and r. f. werner for critical comments .
px acknowledges financial support from the national natural science foundation of china under grant nos 11004029 and 11174052 , the natural science foundation of jiangsu province under grant no bk2010422 , the ph.d.program of the ministry of education of china , the excellent young teachers program of southeast university and the national basic research development program of china ( 973 program ) under grant no 2011cb921203 .
bcs acknowledges financial support from cifar , nserc and aitf .
y. aharonov , l. davidovich and n. zagury , * 48 * , 1687 ( 1993 ) .
d. aharonov , a. ambainis , j. kempe and u. vazirani , proc .
33rd stoc ( new york ) pp 50 - 59 ( 2001 ) .
v. kendon , phil .
a * 364 * , 3407 ( 2006 ) .
t d mackay , s d bartlett , l t stephenson and b c sanders , j. phys .
gen . * 35 * , 2745 ( 2002 ) . c. di franco , m. mc gettrick , and th .
busch , , 080502 ( 2011 ) . c. di franco , m. mc gettrick , t. machida , and th .
busch , , 042337 ( 2011 ) .
t. brun , h. a. carteret and a. ambainis , phys .
a * 67 * , 052317 ( 2003 ) .
t. brun , h. a. carteret and a. ambainis , phys .
* 91 * , 130602 ( 2003 ) .
m. stefanak , s. m. barnett , b. kollar , t. kiss and i. jex , new j. phys . * 13 * , 033029 ( 2011 ) .
p. p. rohde , a. schreiber , m. stefanak , i. jex and c. silberhorn , new j. phys .
* 13 * , 013001 ( 2001 ) .
a. ahlbrecht , h. vogts , a. h. werner and r. f. werner , j. math . phys . * 52 * , 042201 ( 2011 ) .
a. ahlbrecht , a. alberti , d. meschede , v. b. scholz , a. h. werner and r. f. werner , preprint : arxiv:1105.1051 . s. d. berry and j. b. wang , , 042317 ( 2011 )
. m. mohseni , p. rebentrost , s. lloyd and a. aspuru - guzik , j. chem .
* 129 * , 174106 ( 2008 ) .
b. c. travaglione and g. j. milburn , phys
a * 65 * , 032310 ( 2002 ) .
p. xue and b. c. sanders , new j. phys .
* 8 * , 053025 ( 2008 ) .
p. xue , b. c. sanders , a. blais and k. lalumire , phys .
a * 78 * , 042334 ( 2008 ) .
f. helmer , m. mariantoni , a. g. fowler , j. von delft , e. solano and f. marquardt , euro .
* 85 * , 50007 ( 2009 ) .
r. raussendorf , v. m. kendon and h. j. briegel , phys .
a * 66 * , 052319 ( 2002 ) .
j. kempe , contemp .
* 44 * , 307 ( 2003 ) ; n. shenvi , j. kempe and k. b. whaley , phys .
a * 67 * , 052307 ( 2003 ) .
j. p. keating , n. linden , j. c. f. matthews and a. winter , , 012315 ( 2007 ) .
a. schreiber , k. n. cassemiro , v. potoek , a. gbris , i. jex , and ch .
silberhorn , , 180403 ( 2011 ) .
h. ollivier and w. h. zurek , phys .
lett . * 88 * , 017901 ( 2001 ) .
s. luo , * 77 * , 022301 ( 2008 ) .
n. li and s. luo , phys .
a * 76 * , 032327 ( 2007 ) .
s. luo , phys .
a * 77 * , 042303 ( 2008 ) .
s. luo , phys .
a * 73 * , 022324 ( 2006 ) .
v. vedral , * 90 * , 050401 ( 2003 ) .
k. modi , t. paterek , w. son , v. vidal and m. williamson , * 104 * , 080501 ( 2010 ) . o.
giraud , p. braun and d. braun , new j. phys .
* 20 * , 063005 ( 2010 ) .
r. srikanth , s. banerjee and c. m. chandrashekar , * 81 * , 062123 ( 2010 ) .
b. r. rao , r. srikanth , c. m. chandrashekar and s. banerjee , * 83 * , 064302 ( 2011 ) .
d. cavalcanti , l. aolita , s. boixo , k. modi , m. piani and a. winter , * 83 * , 032324 ( 2011 ) . | we consider two independent quantum walks on separate lines augmented by partial or full swapping of coins after each step . for classical random walks , swapping or not swapping coins makes little difference to the random walk characteristics , but we show that quantum walks with partial swapping of coins have complicated yet elegant inter - walker correlations .
specifically we study the joint position distribution of the reduced two - walker state after tracing out the coins and analyze total , classical and quantum correlations in terms of mutual information , quantum mutual information , and measurement - induced disturbance .
our analysis shows intriguing quantum features without classical analogues . |
galactic open clusters are excellent tools to study the evolution of our galaxy from the chemical and structural points of view .
they provide information on the chemical abundances in the disk and relative radial gradients ( e.g. janes 1979 , panagia and tosi 1981 , friel and janes 1993 ) , on the interactions between thin and thick disks ( e.g. sandage 1988 ) , on the average radial velocities and stellar ages at different galactic locations ( e.g. janes and phelps 1994 ) , on the absolute age of the disk .
this latter could also be obtained from isolated field stars , e.g. studying the white dwarfs luminosity function at its cool end ; the actual value is still uncertain , varying from about 6.5 to 13 gyr due to different models for the white dwarfs cores and to different treatments of the cooling and crystallization processes ( see for example winget et al .
1987 , isern et al .
1995 , wood 1995 ) , but the oldest ages ( more than about 9 gyr ) seem to be preferred .
this would imply some kind of temporal continuity between the formation of disk and halo , since the youngest halo clusters are only slightly older than this age ( see e.g. buonanno et al . 1994 , or chaboyer at al . 1996 ) . besides this method
, we are able to reliably date only star clusters , and open clusters represent the only class of objects covering both a large range of distances ( several kpc around the sun ) and a large range of ages ( from a few myr up to @xmath210 gyr ) and can therefore provide key constraints to galactic evolution theories ( e.g. tosi 1995 ) . to this aim
, it is however crucial that the observational data be very accurate and homogeneously treated to avoid misleading effects ( see also carraro and chiosi 1994 ) . in order to study in detail the metallicity and age distribution of open clusters with galactocentric distance ,
we have undertaken a project to analyse with the required accuracy a sample of open clusters located at different galactic radii and supposed to have different ages and metallicities .
deep ccd photometry is taken and properly analysed for each of the examined clusters .
age , metallicity , reddening and distance modulus are derived from the resulting colour - magnitude diagrams ( cmds ) and luminosity functions ( lfs ) through the comparison with the corresponding cmds and lfs generated by a numerical code for montecarlo simulations based on stellar evolution tracks and taking into account theoretical and observational uncertainties ( tosi et al . 1991 ) .
these simulations have proved to be much more powerful than the classical isochrone fitting method to study the evolutionary status of the analysed region both in galactic open clusters ( bonifazi et al . 1990 ) and in nearby irregular galaxies ( marconi et al .
1995 ) . as an interesting by - product of our method
we can evaluate the effects connected to the adoption of different stellar evolution models .
so far we have presented the results on the young metal rich cluster ngc 7790 ( romeo et al . 1989 ) and the old metal poor cluster ngc 2243 ( bonifazi et al .
1990 ) and will shortly present results on the old metal poor clusters ngc 2506 and ngc 6253 and the young cluster ngc 6603 .
the galactic cluster collinder 261 ( cr 261 , c1234@xmath0682 : @xmath3 ) has been found old by phelps et al .
( 1994 ) who find it to be at least as old as ngc 6791 .
friel et al . ( 1995 ) consider it among the oldest open clusters and derive from moderate resolution spectroscopy a metallicity [ fe / h][email protected] .
on the other hand , cr 261 has been found old but metal rich by the recent studies of kaluzny et al .
( 1995 ) and mazur et al .
( 1995 , hereinafter mkk ) . here
we present deep ccd photometry of the cluster in the ubvi bands , from which we derive our own estimates of age , metallicity , distance and reddening .
in section 2 we present the observations and data reductions , in section 3 we introduce the obtained cmds , in section 4 we address the cluster parameters obtained by simulations based on three different classes of stellar models .
the results are discussed in section 5 in the context of structure and evolution of the galactic disk .
[ cols="^ , < , < , < , < , < , < " , ]
figures [ fig-3 ] and [ fig-4 ] show the cmds obtained from our reductions .
as can be seen from figure [ fig-3 ] , the cluster is well visible even when all field stars are plotted . the main sequence to is at v=16.7 , = 0.85 , = 0.95 .
these values are in perfect agreement with those deduced from the only published calibrated cmds ; namely , fig.4 of mkk who have bvi photometry , and fig.24 of phelphs et al .
( 1994 ) , who only have vi . our sequence is better defined , being somewhat less dispersed , and is clearly distinguishable down to about 4 magnitudes fainter than the to .
a few red bright stars are visible in both cmds ( around v=14 , and = 1.45 or = 1.4 ) , and , as done by phelps et al .
( 1994 ) and mkk , we assign them to the red giant clump , corresponding to the core - he burning phase .
the magnitude distance from the red clump to the to is therefore @xmath4v@xmath5 2.7 .
this large value and the structure of the cmd indicate that cr 261 is an old cluster .
mkk show ( their fig.6 ) a few very bright and very red stars : we identified them in our frames , but they were saturated even in the shortest exposures .
as already said , we have u measurements only for the central field .
the classical b@xmath0v @xmath6 u@xmath0b plane could not be used to determine the reddening , since the to stars are too cold , and anyway our u calibration is not of the best quality .
figure [ fig-4 ] presents the cmds involving the u band ; in all three cmds the ms is well defined , and we can clearly see a lot of blue straggler stars .
furthermore , the subgiant / red giant branch and the red clump are quite apparent . looking in particular at the u @xmath6 ( u v ) diagram ,
a hint of binary sequence may be seen : it lies just about 0.7 mag brighter than the ms ridge line .
the other two cmds are not so useful in this case . to better assess the significance of this binary sequence we built histograms in colour at different magnitude levels : unfortunately only in one interval a secondary peak due to the possible binaries can be clearly seen .
this is not enough to rule out the possibility of a binary sequence ; furthermore , mkk found a large number of eclipsing binary systems in cr 261 , and several of them lie in the right region of the cmd ( see their fig.11 ) .
deconvolution of the cluster from the field stars is quite difficult , if not flatly impossible , since the latter dominate : they much outnumber the cluster stars even in the central part ( radius about 3.5 arcmin ) . unfortunately , cr 261 was given in the literature a much smaller radius ( 5 arcmin diameter , lang 1992 ) than it actually has , so even our nominal external field contains several cluster members .
the same problem was faced by mkk who observed a region of 16.5 by 16.5 arcmin centered on cr 261 , but were not able to define a convincing external control field for decontamination , and concluded that the cluster radius must be at least 6 arcmin .
we decided to build an artificial `` external field '' , using the 4 dutch and the easternmost danish fields .
of them we retained only the parts lying more than 12 arcmin from the cluster centre .
we then scaled the number of stars to be taken into account for decontamination with the ratio of the areas : since the region from which we extracted the field stars is smaller than the central field we simulated the required number of stars , varying magnitudes and colours of the real ones following gaussian random distributions within the ranges of the empirical photometric errors .
the resulting cluster / field stars ratio in the central area of radius 3.5@xmath7 is about 0.25 which means that the actual cluster members of that area are approximately 800 .
figure [ fig-5 ] ( a ) shows the cmd of the stars detected in the field between 12 and 13 arcmin from the centre and figure [ fig-5 ] ( b ) the semi - empirical diagram resulting from the above procedure to scale the observed stars to the area of the central region of 3.5@xmath7 . due to the uncertainty of such procedure , we do not attempt any decontamination of the cluster field : as will be seen in the next session , we consider it safer to add field stars to the simulations rather than cleaning the cluster cmd from them .
in their detailed study of binary stars in cr 261 , mkk suggest for this cluster an age of @xmath88 gyr , a distance modulus around 13 mag and a reddening = 0.22 , all based on the similarity between the features of its cmd and those of other old open clusters , in particular of be 39 .
in addition , the discovery of extremely red giants with @xmath93 has led these authors to include cr 261 among the clusters with solar or higher metallicity ( kaluzny et al .
this suggestion is however in contrast with the [ fe / h][email protected] inferred by friel et al .
( 1995 ) from spectroscopic indices . here ,
we follow a different procedure to estimate the values of these parameters from our own photometry .
we have applied to cr 261 the method described by tosi et al .
( 1991 ) for nearby irregular galaxies , which is an expansion of the classical isochrone fitting and allows to simultaneously derive the age , reddening and distance modulus of the cluster .
the method compares the observed cmds of the system with synthetic cmds resulting from monte carlo simulations using the same number of stars above the same limiting magnitude , and with the same photometric errors and incompleteness factors in each magnitude bin as derived from the photometric data .
the theoretical diagrams are transformed into the empirical ( v , ) plane by finding the values of reddening and distance modulus providing the best agreement with the stellar distribution in the observed cmd .
if a synthetic cmd coincides with the observed one , it thus yields the best age , reddening and distance modulus pertaining to the cluster in the framework of the adopted stellar evolution models .
the resulting set of values , however , is not unique , because different stellar evolution models may produce different solutions .
for this reason , we have derived the cluster parameters with three different data bases of stellar models , already known to predict rather different ages : this approach allows us to evaluate both the best parameter values and the corresponding theoretical uncertainties .
the synthetic diagrams examined for cr 261 are based on homogeneous sets of stellar evolution tracks computed for several initial metallicities and already proven by their authors to reproduce many observational constraints .
our simulations have been performed with : a ) the tracks with classical mixing length treatment of the convective zones computed by castellani et al .
( 1993 , hereinafter franec ) , b ) the tracks with overshooting from convective cores by the padova group ( hereinafter bbc ) , c ) the tracks by dantona et al .
( 1992 , hereinafter cm ) , with the new convection treatment proposed by canuto & mazzitelli ( 1991 ) . within the framework of each group of stellar models
, we have performed several simulations for any reasonable combination of age , reddening and distance modulus , all of which have been compared with the empirical cmd and luminosity functions of cr 261 .
we describe below only the best cases , selected on the basis of these comparisons .
the best cases for each of the three classes of stellar models are shown in figure [ fig-6 ] ; other cases , useful for a better understanding of the cluster conditions , are shown in figure [ fig-7 ] .
due to the large back / foreground contamination affecting cr261 , and for an easier comparison with mkk s diagrams , we have limited our analysis to the stars located within 3.5@xmath7 from the cluster center .
this selection does not introduce any bias in the derived parameter values , since the cmd of different regions of cr 261 are totally equivalent to each other ( see figure [ fig-3 ] ) .
the contamination in this central area is less severe , but unfortunately it still prevents to derive detailed information on some of the cluster characteristics , such as the colour and luminosity functions of the redder stars and the relative number of stars in the various evolutionary phases . of the 11243 stars detected in b ,
v and i bands with @xmath100.1 mag , 2779 lie within a radius of 3.5@xmath7 .
the cmd of this selected sample is shown in panel ( g ) of figure [ fig-6 ] .
unfortunately the majority of the objects are probably field stars and ( see discussion in section 3 ) only about 800 can be safely considered as cluster members .
the synthetic diagrams discussed below therefore assume the cluster to be populated by 800 stars .
however , for further test , we have also performed several montecarlo simulations based on higher membership and found that an adopted population of 1000 stars , or more , definitely provides cmds with excessively crowded main sequence and to regions , when compared with the observational diagram .
this means that our estimate of 800 cluster members situated in the inner 3.5@xmath7 , based on a statistical subtraction of the external field , is consistent with the number of stars appearing to populate both the ms and post - ms regions of the cmd .
we have not included binary stars in our simulations , because the field contamination is too large to allow to derive any significant information on their cluster distribution from the observed cmds of cr 261 .
we are however inclined to believe that the fraction of binaries is fairly large in this system , first because of the large number of blue stragglers , second because of the spread in the ms and post - ms stellar distribution which is much larger than expected from the photometric error , and third because of the relatively large number , 45 , of binary stars actually discovered by mkk .
franec sets of models follow the evolution of stars between 0.6 and 1 m@xmath11 in the central hydrogen burning phases and the evolution of stars between 1 and 9 m@xmath11 up to the onset of thermal pulses on the asymptotic giant branch .
they have been computed for several values of initial helium and metal abundances : for the case of cr 261 , we have used the tracks with y and z equal to ( 0.27 , 0.01 ) , ( 0.27 , 0.02 ) , ( 0.30 , 0.02 ) .
however , all the synthetic diagrams in better agreement with the data are found to be based on the set with composition ( 0.27 , 0.02 ) . models with higher helium content
in fact present a fairly brighter clump during the core helium burning phase , which implies a larger magnitude difference between clump and to stars , inconsistent with that derived from the observational data .
models with metallicity lower than solar present a rounder shape of the to region and a more extended subgiant branch ( see the best case in figure [ fig-7 ] ( c ) ) , which make more difficult a good reproduction of the observed diagram . figure [ fig-6 ] ( a ) shows the cmd assuming @xmath12=7 gyr , ( m - m)@xmath1=11.9 and = 0.30 .
the cluster luminosity , colour and stellar distribution in the ms and to regions are well reproduced by this model , as can be more easily appreciated in figure [ fig-6 ] ( d ) where we have superimposed to the synthetic diagram the semi - empirical cmd of the external field , discussed in section 3 .
stars in the post - ms phases appear more concentrated in the synthetic cmd than in the observational one .
nonetheless , their distribution is consistent with the data if one considers that cr 261 is probably populated by a significant fraction of binary stars whose interactions during the giant evolutionary phases may well be the cause of the observed spread in the cmd . the same good result is obtained assuming @xmath12=8 gyr , ( m - m)@xmath1=11.8 and = 0.30 , and the two synthetic cmds are in practice indistinguishable .
notice however that , since age , distance modulus and reddening all affect the apparent magnitude of the stars , to reproduce the observational diagram of cr261 with an age higher than in the previous case , we must assume a slightly smaller distance modulus to compensate the correspondingly lower intrinsic luminosity of the to .
models with @xmath138 gyr are less satisfactory , because to keep on fitting the ms and to regions one needs to assign to the cluster reddenings or distance moduli which make the synthetic red giant branch ( rgb ) and clump definitely bluer than observed , as shown in figure [ fig-7 ] ( b ) , where the cmd corresponding to 10 gyr , = 0.27 and ( m - m)@xmath1=11.7 is shown . in other words , for ages older than 8 gyr , we do not find a combination of and ( m - m)@xmath1 able to well reproduce the cmd features .
an age slightly younger than 7 gyr might also be attributable to cr 261 , with the same reddening and slightly larger distance .
for instance , the model with @xmath12=6 gyr , ( m - m)@xmath1=12.0 and = 0.30 , shown in figure [ fig-7 ] ( a ) still reproduces most of the observational features of cr261 . however , at @xmath12=6 gyr the gap corresponding to the ms overall contraction phase , typical of clusters some gyr old ( e.g. ngc 2243 , see bonifazi et al .
1990 ) , is present in the synthetic cmd . since the gap does not appear in the observational diagram , cr 261 must be older than 6 gyr .
notice that , had we used the classical isochrone fitting method to infer the cluster parameters , we would not have noticed the gap and would not have been able to put this lower limit to the age of cr 261 . for even younger ages the shapes of both the ms and the to regions change , with the ms becoming more straight and the to showing the hook typical of young systems .
these modifications make such younger diagrams inconsistent with the data . in summary , adopting franec evolutionary tracks , we find that cr 261 has an age between 7 and 8 gyr , a distance modulus in the range 11.8 - 11.9 and a reddening 0.3 , and is consistent with an initial chemical composition nominally solar but actually lower ( see note ) , and consistent with the value [ fe / h]@xmath140.14 derived by friel et al .
( 1995 ) from low dispersion spectra .
the stellar evolution tracks computed by the padova group take into account the possible overshooting of convective regions out of the edges defined by the classical mixing length theories .
they have been computed for the whole mass range ( i.e. between 0.5 - 0.8 and 100 - 120 m@xmath11 ) and several initial metallicities .
they reach the tip of the asymptotic giant branch or the ignition of the core c - o burning phase , depending on the initial stellar mass . for cr 261
, we have tested the sets of tracks with y and z equal to ( 0.28 , 0.008 ) by alongi et al .
( 1993 ) , ( 0.28 , 0.02 ) by bressan et al .
( 1994 ) and ( 0.352 , 0.05 ) by fagotto et al .
( 1994 ) , available at the strasbourg data center . figure [ fig-6 ] ( b ) shows one of the bbc synthetic cmds in better agreement with the data .
it assumes solar chemical composition , @xmath12=11 gyr , = 0.25 and ( m - m)@xmath1=11.7 and reproduces pretty well both the ms and post - ms colours , luminosities and stellar distributions , as can be better appreciated in panel ( e ) where the synthetic cmd overlaps that of the semi - empirical external field .
equivalent results are obtained with @xmath12=9 gyr , = 0.27 and ( m - m)@xmath1=11.7 and with @xmath12=10 gyr , = 0.26 and ( m - m)@xmath1=11.7 . however
, 9 gyr is the minimum age attributable to cr 261 with this set of tracks , because at smaller @xmath12 they show a large ms gap and a hooked to , inconsistent with the observed shape , as already evident in the cmd in figure [ fig-7 ] ( d ) , where the adopted parameters are @xmath12=8 gyr , = 0.3 and ( m - m)@xmath1=11.8 . on the other hand
, the cluster might be as old as 12 gyr , but the shape of the ms and post - ms phases at this age starts to deviate from the observed ones to become more similar to those observed in globular clusters .
for instance , the subgiant branch runs at fairly constant luminosity and the base of the rgb is increasingly brighter than observed . the same age ( 10@xmath151 gyr ) and
distance modulus ( 11.7 ) are obtained also using the more metal rich set of tracks .
a significantly lower reddening 0.16@[email protected] is however required to compensate the redder intrinsic colours due to the higher metallicity .
the problem with this set of tracks is that , probably because of the large helium content , the predicted red giant clump is more luminous than with the solar composition set , and 0.5 - 1 magnitude brighter than observed in cr 261 ( see e.g. figure [ fig-7 ] ( e ) for @xmath12=9 gyr , = 0.19 and ( m - m)@xmath1=11.7 ) an age of 9 - 10 gyr and the same distance modulus 11.7 are inferred also with the lower metallicity set of models ( z=0.008 ) . since these tracks are intrinsically bluer , because of the lower metal content , they require a slightly larger reddening = 0.32 - 0.34 to fit the data .
figure [ fig-7 ] ( f ) shows the synthetic cmd with @xmath12=9 gyr , = 0.34 and ( m - m)=11.7 . in these stellar tracks ,
the rgb spans a larger colour range than in any other of the evolutionary sets considered here and this leads to the shallower slope of the giant branch visible in figure [ fig-7 ] ( f ) .
given the spread in the corresponding observational distribution , we are however unable to evaluate what is the appropriate slope . instead
, we consider the comparison between this low metallicity tracks and the observational diagram less satisfactory than that of the solar metallicity models because of the shape of the to region and the larger colour extension of the subgiant branch .
therefore , we can conclude that with bbc models the age of cr 261 is @xmath1711 gyr and the distance modulus ( m - m)@xmath1=11.7 , independently of the cluster metallicity .
the solar composition set of tracks seems the most appropriate and implies a reddening [email protected] .
a few years ago , canuto & mazzitelli ( 1991 ) have proposed a new approach to the treatment of the stellar convective zones , alternative to the classic mixing length theory , and dantona et al .
( 1992 ) have computed with such approach the evolutionary tracks of stars of initial mass between 0.65 and 1.5 m@xmath11 and chemical abundance y=0.285 and z=0.018 , up to the core helium burning .
these tracks show interesting differences in the relative timescales , luminosities and temperatures of the various phases and we thus consider it useful to compare their predictions with the observed cmd of cr 261
. figure [ fig-6 ] ( c ) and ( f ) show the synthetic cmd based on the cm tracks in better agreement with our observational diagram .
it assumes an age of 6 gyr , ( m - m)@xmath1=12.0 and = 0.34 .
similar results are obtained with @xmath12=7 gyr if the lower to luminosity is compensated by a lower reddening .
these tracks show a curvature of the ms just below the to higher than with classic models and a slightly hotter luminosity minimum at the base of the red giant branch .
unfortunately the field contamination is too large to allow any choice between the cm and classical cases and , within the uncertainties , both look in very good agreement with the data .
comparing observational data with the synthetic cmds is useful not only to measure fundamental quantities for the observed objects , but also to better define the actual uncertainties associated with stellar evolutionary models .
in fact , despite the very different assumptions of the three sets of evolutionary tracks adopted here , and the different solutions obtained from each of them , we are unable to univocally and uncontroversially choose @xmath18 best fit among them : considering the observational errors all three sets have at least one good solution for cr 261 .
this clearly translates into a `` theoretical uncertainty '' on the derived parameters , which affects most strongly the age .
also , this implies that age estimates from a single set of isochrones ( beyond the fact that isochrone fitting is a less powerful tool than comparison to synthetic cmds ) hide a much larger uncertainty than quoted , since the latter only reflects the @xmath19 error for the chosen set of models .
finally , this also means than no ranking in age can be truly believed if it has not been derived in a homogeneous way , as already remarked by e.g. carraro & chiosi ( 1994 ) and friel ( 1995 ) .
taking into account the mentioned uncertainties , we prefer to give for cr 261 not a simple `` best value '' for the quantities , but instead the following ranges : 7@xmath2011 gyr , 0.25@xmath5e(b - v)@xmath50.34 , 11.7@xmath5(m - m)@xmath2112.0 .
metallicity is about solar , with indication for being slightly lower . how do our findings compare to the literature on cr 261 ?
phelps et al .
( 1994 ) give a @xmath4v ( difference in magnitude between red clump and to ) of 2.6 ( in agreement with our @xmath52.7 ) , second only to be 17 , and equal to that attributed to ngc 6791 , for which the age estimates vary from about 5.5 to 10 gyr .
an age ranking based on @xmath4v is not strictly monotonic and may be rather unsafe in cases like cr 261 where the red clump is poorly defined , but such large @xmath4v undoubtedly indicates very old ages .
phelps et al .
are inclined towards the upper end of the age range and quote the 9 gyr assigned to ngc 6791 by garnavich et al .
( 1994 ) when discussing the age difference between the oldest open clusters and the younger globular clusters .
they do not give any indication about metallicity or reddening .
our age estimate is in good agreement with their suggestion : cr 261 is surely among the oldest open clusters of our galaxy .
mkk measure a @xmath4v of 2.4 - 2.5 , comparable to that of ngc 188 , implying a slightly younger age : their estimate is of about 6 - 8 gyr .
they find distance modulus and reddening , ( m - m)@xmath1=13.0 , = 0.22 , from comparison to be 39 , but admit that this is only a preliminary determination .
their estimate of metallicity ( comparable to solar or higher ) is based on the presence of very red giants , by analogy to ngc 6791 .
our age estimate is consistent with mkk s , but and distance modulus are not .
we regard our estimate as more accurate than theirs , since it was explicitly beyond the scope of their work to determine precise values for those quantities , and they only do it by comparison with other clusters and one single isochrone .
our values are systematically different from theirs , independently of the adopted set of evolutionary tracks . in other words , _ none _ of the synthetic cmds is able to reproduce the observational one assuming a distance modulus as large as 13 .
the difference with mkk parameters is more significant if one considers that a reddening as low as their 0.22 is acceptable in our simulations only if we adopt the bbc high metallicity tracks ( morphologically inconsistent with the observed cmd ) and ( m - m)@xmath1=11.7 , i.e. much lower than their suggested 13 .
vice versa , our reddenning and distance modulus ranges are perfectly consistent with the values , 0.33 and 12.04 respectively , derived by janes & phelps ( 1994 ) .
concerning the chemical composition , we did not find any indication in favour of a metallicity higher than solar . on the contrary ,
our best fits have always been for solar metallicity ; furthermore , as noted before , for the franec tracks the nominally - solar set of models corresponds to a slightly lower effective metallicity .
this is in good agreement with the only metallicity spectroscopically determined ( [ fe / h]=0.14 , friel et al .
1995 ) .
whatever its age , cr 261 is clearly a disk object : its galactic coordinates , combined with the distance from the sun derived from the distance modulus , place it just in close proximity of the sagittarius spiral arm .
what we are seeing now is probably the surviving portion of a much bigger stellar system which has had the time to spread around many but not all of its original members , as suggested by the large spatial extension , but with very moderate concentration , of the region containing stars with cr 261 features .
its very old age ( greater than 6 gyr , and possibly as high as 11 gyr ) , close to that of the youngest globulars , and the fact that it does not seem to be the oldest known open cluster , indicate that the disk contains an old population of relatively metal rich objects that must be accounted for by realistic galactic formation and evolution models , both dynamical and chemical .
we wish to thank laura greggio for the code which has been the bulk of the numerical procedure for cmd simulations .
we are grateful to franca dantona , sandro bressan and oscar straniero for useful conversations on the evolutionary tracks .
this research has made use of the simbad database , operated at cds , strasbourg , france . | we present ubvi photometry for the old open cluster collinder 261 . from comparison of the observed colour - magnitude diagrams with simulations based on stellar evolutionary models we derive in a self consistent way reddening , distance , and age of the cluster : = [email protected] , ( m - m)@xmath1=11.7 - 12.0 , and age=7@xmath08 gyr or 9@xmath011 gyr , depending on the adopted stellar tracks .
the models in better agreement with the data turn out to have metallicity at most solar .
= -1 cm hertzsprung - russel ( hr ) diagram open clusters and associations : individual : collinder 261 age metallicity |
over one hundred years ago , the italian economist vilfredo pareto @xcite made one of the first empirical studies of the distribution of wealth by undertaking a careful study of land ownership in italy , switzerland and germany .
in the course of this study , he plotted the fraction of economic agents with land holdings worth more than @xmath1 as a function of @xmath1 .
his studies led him to believe that this function , which we shall denote by @xmath2 has a universal form . by its definition , it is clear that @xmath3 is monotone non - increasing , and that @xmath4 and @xmath5 .
what pareto observed is that @xmath2 is approximately equal to one for all @xmath1 less than a certain cutoff value denoted by @xmath6 , and decays as a power law for @xmath7 .
that is , pareto s empirical observations led him to conclude that it is approximately true that @xmath8 the power @xmath9 is usually called the pareto exponent . to put pareto s observations in modern terms , we may note that @xmath10 is the cumulative distribution function ( cdf ) of economic agents , ordered by wealth .
let us denote the corresponding probability density function ( pdf ) of agents by @xmath11 , but we shall adopt the convention of normalizing @xmath12 to the total number of economic agents , rather than to unity , so that @xmath13 is the total number of agents with wealth in @xmath14 $ ] , for @xmath15 .
the total number of economic agents is given by the zeroth moment of the pdf of agents , @xmath16 and the total amount of wealth in the economy is given by the first moment , @xmath17 with the above nomenclature established , the function that pareto plotted is @xmath18 from the fundamental theorem of calculus , we see that the pdf of agents is related to the derivative of pareto s function by exist , at least in the distributional sense . ]
@xmath19 in terms of the pdf of agents , pareto s empirical observation is then @xmath20 the ubiquity of the pareto distribution is indeed a very fundamental observation of macroeconomics , but it has resisted a microeconomic explanation for most of the last century .
recent work @xcite has suggested that the shape of pareto s curve is due to the fact that wealth distributions satisfy a certain boltzmann equation . before describing this boltzmann equation ,
we first examine another common quantification of inequality .
in 1906 , at approximately the same time as pareto was making his observations in europe , the american economist max o. lorenz was completing his dissertation work at the university of wisconsin madison in which he plotted the cumulative distribution of wealth versus the cumulative distribution of agents .
that is , lorenz plotted the quantity @xmath21 versus the quantity @xmath22 both @xmath23 and @xmath24 are monotone nondecreasing , and they increase from zero to one as @xmath1 goes from zero to infinity .
this graph , which has come to be called the _ lorenz curve _ , can be plotted in the unit square , as shown schematically in fig .
[ fig : pareto - lorenz ] . is plotted on the abscissa , while that of wealth @xmath23 is plotted on the ordinate .
the parameter @xmath1 ranges from zero in the lower left corner of the figure , to infinity at the upper right .
the diagonal blue line connecting those two corners corresponds to complete economic equality , in which all agents have the same amount of wealth .
the orange curve is a more realistic depiction of the distribution of wealth in a society .
the fraction of the triangular area under the diagonal that is shaded in blue is the gini coefficient . ]
if all wealth in a society were equally distributed , the lorenz curve would be the diagonal line , plotted in blue in fig .
[ fig : pareto - lorenz ] .
this is because any fraction @xmath25 $ ] of the agents would possess the same fraction @xmath26 of the total wealth . in real economies
, this curve always lies below the diagonal , more like the orange curve in fig .
[ fig : pareto - lorenz ] . to prove that the lorenz curve can never pass above the diagonal ,
note first that @xmath27 so the slope of the lorenz curve is given by @xmath28 which is the ratio of the wealth @xmath1 corresponding to that point on the curve to the average wealth .
it follows that @xmath29 where we have abused notation , in that @xmath30 is now considered a function of @xmath31 .
we see that @xmath32 when @xmath33 , and that @xmath34 when @xmath35 , and moreover that @xmath36 is concave up .
it follows that @xmath36 is bounded above by the diagonal .
equations ( [ eq : dl ] ) and ( [ eq : df ] ) relate the quantities plotted in the lorenz curve to @xmath11 , and we may also note that @xmath37 thus , given @xmath38 and @xmath39 , we see that any of the quantities @xmath2 , @xmath11 , @xmath23 and @xmath24 can be derived from any other , so they all contain equivalent information which is to say that they all contain essentially complete information about the distribution of wealth in a society .
in fact , actual data for wealth distributions in the world today is very scant , and economists have to content themselves with much coarser characterizations of wealth inequality than the quantities described above .
one of the most popular of these is due to the italian statistician and sociologist corrado gini , who also worked roughly contemporaneously with pareto and lorenz . the _ gini coefficient _
@xmath40 is defined as the ratio of the shaded area in fig .
[ fig : pareto - lorenz ] , lying between the diagonal segment and the lorenz curve .
consequently , @xmath41 when everybody has equal wealth ; the limit @xmath42 describes the approach to complete oligarchy . in terms of the quantities defined above , the gini coefficient is given by @xmath43 using eqs .
( [ eq : dl ] ) and ( [ eq : df ] ) , we can change integration variables from @xmath31 to @xmath1 to find @xmath44 alternatively , changing the order of integration yields @xmath45 where we have changed the remaining variable of integration from @xmath46 to @xmath1 in the rightmost expression .
equation ( [ eq : gini1 ] ) indicates that @xmath47 , where the angle brackets denote an average over the pdf @xmath12 .
likewise , eq . ( [ eq : gini2 ] ) indicates that @xmath48 . from these fundamental relationships between the gini coefficient and the functions introduced earlier , we see that @xmath40 is a quadratic functional of @xmath12 , and we may sometimes emphasize this functional dependence by writing it as @xmath49 $ ] . in what follows , we shall need the _ frchet derivative _ of @xmath49 $ ] with respect to @xmath12 , which is the analog of the gradient in function space .
this is defined by noting that , for any sufficiently well behaved function @xmath50 , @xmath51 - g[p ] \right)\ ] ] is a linear functional of @xmath52 . as such , by the riesz representation theorem , it can be expressed as the inner product of @xmath52 and a quantity which we shall denote by @xmath53 ; that is , for all sufficiently well behaved functions @xmath52 , the relation @xmath54 - g[p ] \right)\ ] ] defines the frchet derivative , @xmath53 , where the parentheses on the left indicate the standard @xmath55 inner product .
this may be thought of as the infinite - dimensional version of @xmath56 for a scalar function @xmath57 .
applying the above definition to eq .
( [ eq : gini2 ] ) , it is a straightforward calculation to find @xmath58 .
\label{eq : frechet}\ ] ] now , if the pdf , @xmath12 , changes in time , it will cause a change in the gini coefficient given by @xmath59 which is an infinite - dimensional version of the ordinary chain rule , @xmath60 in the next two sections , we shall consider two different dynamical equations for @xmath12 , and use them along with eq .
( [ eq : dgdt ] ) to place bounds on @xmath61 .
as noted in the introduction , though pareto s law is over a century old , an explanation for it in terms of microeconomic exchange relations between individual agents is still elusive .
the general idea that simple rules for asset exchange might be used to explain wealth distributions appears to be due to angle @xcite in 1986 .
such models have come to be called _ asset - exchange models _ ( aems ) and they typically involve binary transactions between agents involving some increment of wealth @xmath62 , with rules for which agent gains it and which agent loses it . the first work applying the mathematical methods of kinetic theory to such models is the paper of ispolatov , krapivsky and redner @xcite in 1998 .
they considered an aem model in which the agent who loses the wealth is selected with even odds , and in which @xmath62 was proportional to the wealth of the losing agent .
writing in a popular article in 2002 , hayes @xcite noted that in an economy governed by this model , no agent would willingly trade with a poorer agent , and therefore would try to use deception to trade only with wealthier agents .
for this reason , he named the model of ispolatov et al . the _ theft - and - fraud model _
( tfm ) . in the same article
mentioned above , hayes @xcite proposed a variant of the tfm , in which the losing agent is still selected with even odds , but in which @xmath62 is proportional to the wealth of the poorer agent , rather than that of the losing agent .
he referred to this as the _ yard - sale model _
( ysm ) and noted that it describes an economy in which rich and poor will have no strategic reason not to trade with each other .
the first kinetic theoretical description of the ysm was given by boghosian in 2014 @xcite .
in analogy with @xcite , he derived a boltzmann equation for the ysm , but then went on to show that in the limit of small @xmath62 and frequent transactions , this reduced to a certain nonlinear , nonlocal fokker - planck equation .
he presented numerical evidence indicating that the ysm by itself exhibits `` wealth condensation , '' in which all the wealth ends up in the hands of a single oligarch .
he also showed that , when supplemented with a simple model for redistribution , the ysm yields pareto - like wealth distributions , including a cutoff at low wealth and an approximate power law at large wealth , very reminiscent of eq .
( [ eq : parteopdf ] ) .
the ysm can be described by a very simple algorithm .
the version of the algorithm we shall use here is completely equivalent to that described by boghosian @xcite , though we state it in the following slightly different fashion : choose two agents from the population at random to engage in a financial transaction .
call them agent 1 and agent 2 , and denote their respective wealth values by @xmath63 and @xmath64 .
the amount of wealth that will be transferred from agent 1 to agent 2 in this transaction is then @xmath65 , where @xmath66 is sampled from a _ symmetric _ probability density function denoted by @xmath67 .
note that , because @xmath67 is symmetric , the two agents both have even odds of winning and losing .
it may seem odd that an algorithm which gives even odds of winning to both agents engaging in a transaction would cause wealth to concentrate , but this is indeed the case .
this was demonstrated by extensive numerical simulations in @xcite , and there it was conjectured that the time - asymptotic state of @xmath11 is a generalized function , with support at @xmath33 , zeroth and first moments given by @xmath38 and @xmath39 respectively , and possibly divergent higher moments .
this corresponds to an oligarchical state with @xmath68 . in this paper
, we confirm that conjecture by demonstrating that the gini coefficient is a monotone increasing lyapunov functional of the boltzmann equation for the ysm , and that it reaches a maximum value of @xmath68 in the above - described oligarchical state .
the boltzmann equation of the yard - sale model of asset exchange may be written @xmath69 + \frac{1}{n}\int_0^{\frac{w}{1+\beta}}dx \ ; p(x ) \left[p(w-\beta x)-\frac{1}{1+\beta}p\left(\frac{w}{1+\beta}\right)\right ] \right\ } , \label{eq : boltzmann}\ ] ] where @xmath67 is the symmetric distribution described in the last section .
this was derived by boghosian @xcite using arguments similar to the derivation of the molecular boltzmann equation and also using a master equation approach , and so we shall not re - derive it in this paper . instead
, we shall demonstrate that the gini coefficient is a lyapunov function for this equation .
that is , we shall show that @xmath40 is monotone non - decreasing as a consequence of the above dynamics for @xmath12 .
the lyapunov function of the molecular boltzmann equation is traditionally called boltzmann s `` @xmath0 function . ''
adopting that nomenclature , we will show that the gini coefficient is an @xmath0 function for the yard - sale model boltzmann equation .
our first task will be to substitute eq .
( [ eq : boltzmann ] ) into eq .
( [ eq : dgdt ] ) , and demonstrate that @xmath61 thereby obtained is greater than or equal to zero .
we partition this task by rewriting the above as @xmath70 where we have defined @xmath71 , \label{eq : boltzmann1}\ ] ] and @xmath72 .
\label{eq : boltzmann2}\ ] ] we shall now show that the corresponding rates of increase of the gini coefficient given by eq .
( [ eq : dgdt ] ) , @xmath73 are separately greater than or equal to zero for @xmath74 .
we first consider @xmath75 .
combining eqs .
( [ eq : frechet ] ) , ( [ eq : boltzmann1 ] ) and ( [ eq : dgdtj ] ) , we have @xmath76 \int_{-1}^1d\beta \;\eta(\beta ) \left [ p(w)-\frac{1}{1+\beta}p\left(\frac{w}{1+\beta}\right ) \right]\\ & = & \frac{2}{w}\int_{-1}^1d\beta \;\eta(\beta)\int_0^\infty dw\ ; wp(w ) -\frac{2}{w}\int_{-1}^1d\beta \;\eta(\beta)\int_0^\infty dw\ ; w\frac{1}{1+\beta}p\left(\frac{w}{1+\beta}\right)\\ & & + \frac{2}{nw}\int_{-1}^1d\beta \;\eta(\beta ) \int_0^\infty dw\;\frac{1}{1+\beta}p\left(\frac{w}{1+\beta}\right)\int_0^w dx\ ; p(x)(w - x)\\ & & -\frac{2}{nw}\int_{-1}^1d\beta \;\eta(\beta ) \int_0^\infty dw\;p(w)\int_0^w dx\ ; p(x)(w - x).\end{aligned}\ ] ] use the change of variables @xmath77 in the second and third terms , and then change the name of the integration variable from @xmath78 back to @xmath1 to obtain @xmath79.\end{aligned}\ ] ] the first term above vanishes because the integrand is odd in @xmath80 , integrated from @xmath81 to @xmath82 .
the second term is nonnegative because @xmath83 is a non - decreasing function of @xmath1 , as follows from @xmath84 thus we have demonstrated that @xmath85 we next consider @xmath86 . combining eqs .
( [ eq : frechet ] ) , ( [ eq : boltzmann2 ] ) and ( [ eq : dgdtj ] ) , we have @xmath87 \\ & & \int_0^{\frac{w}{1+\beta}}dy\ ; p(y ) \left[p(w-\beta y)-\frac{1}{1+\beta}p\left(\frac{w}{1+\beta}\right)\right]\\ & = & -\frac{2}{nw}\int_{-1}^1d\beta \;\eta(\beta ) \int_0^\infty dw\;w \int_0^{\frac{w}{1+\beta}}dy\ ; p(y)p(w-\beta y ) \\ & & + \frac{2}{nw}\int_{-1}^1d\beta \;\eta(\beta)\frac{1}{1+\beta } \int_0^\infty dw\;w \int_0^{\frac{w}{1+\beta}}dy\ ; p(y)p\left(\frac{w}{1+\beta}\right ) \\ & & + \frac{2}{n^2w}\int_{-1}^1d\beta \;\eta(\beta ) \int_0^\infty dw\ ; \int_0^w dx\ ; p(x)(w - x ) \int_0^{\frac{w}{1+\beta}}dy\ ; p(y)p(w-\beta
y ) \\ & & -\frac{2}{n^2w}\int_{-1}^1d\beta \;\eta(\beta)\frac{1}{1+\beta } \int_0^\infty dw \int_0^w dx\ ; p(x)(w - x ) \int_0^{\frac{w}{1+\beta}}dy\ ; p(y)p\left(\frac{w}{1+\beta}\right).\end{aligned}\ ] ] in the first and third terms , swap the order of integration so that @xmath88 is outermost , and then make the substitution @xmath89 . in the second and fourth terms , use the change of variable @xmath90 , and then swap the order of integration so that @xmath88 is outermost .
the result is @xmath91 the integral over @xmath80 can be performed for the first two terms , whereupon they cancel .
now swap the order of integration in the remaining two integrals so that @xmath92 is outermost , followed by @xmath88 , and then @xmath78 to obtain @xmath93 \\ & & + \frac{2}{n^{2}w } \int_{-1}^{+1 } d\beta \ ; \eta ( \beta ) \int_{0}^{\infty } dx \ ; p(x ) \int_{\frac{x}{1+\beta}}^{\infty } dy \ ;
p(y ) \left [ \beta \int_{y}^{\infty } du \ ; p(u)y - \beta \int_{y}^{\infty } du \ ; p(u)u \right].\end{aligned}\ ] ] in the second term above , note that @xmath94 .
the term with @xmath95 then vanishes upon integration over @xmath80 , leaving us with @xmath96.\end{aligned}\ ] ] next note that because @xmath97 and @xmath98 , it follows that @xmath99 so the integral over @xmath78 from @xmath88 to @xmath100 can be split into one from @xmath88 to @xmath101 plus another from @xmath101 to @xmath100 , resulting in @xmath102\\ & = & + \frac{2}{n^{2}w } \int_{-1}^{+1 } d\beta \ ; \eta(\beta ) \int_{0}^{\infty } dx \ ; p(x ) \int_{0}^{\frac{x}{1+\beta } } dy \ ; p(y)\\ & & \;\;\;\;\;\;\;\;\;\ ; \left [ \int_{\frac{x}{1+\beta}}^{x-\beta y } du \ ; p(u)((x-\beta y)-u ) + \int_{y}^{\frac{x}{1+\beta } } du \
; p(u)(u - y ) \right].\end{aligned}\ ] ] the integrands are now manifestly positive , and we may conclude @xmath103 the above demonstration is admittedly tedious
. there may be a shorter route to this result , but as of this writing we have not been able to find one .
combining eqs .
( [ eq : dgdt1geq0 ] ) and ( [ eq : dgdt2geq0 ] ) , we have @xmath104 and so the gini coefficient is proven to be a lyapunov functional of the boltzmann equation .
in earlier work , boghosian @xcite demonstrated that in the limit of small @xmath80 , the boltzmann equation , eq .
( [ eq : boltzmann ] ) , reduces to a fokker - planck equation of the form @xmath105 where @xmath106 and @xmath107 is a constant . from its method of derivation
, one would expect that @xmath40 is also a lyapunov functional of this equation . here
we demonstrate this directly . combining eqs .
( [ eq : frechet ] ) , ( [ eq : dgdt ] ) and ( [ eq : fp ] ) yields @xmath108 \frac{\partial^2}{\partial w^2}\left(\gamma \frac{w^2}{2}c(w)p(w)\right)\\ & = & -\frac{2}{w}\int_0^\infty dw \ ; \left[-1+\frac{1}{n}\int_0^w dx\ ; p(x)\right ] \frac{\partial}{\partial w}\left(\gamma \frac{w^2}{2}c(w)p(w)\right ) \\ & = & \frac{2}{w}\int_0^\infty dw \ ; \frac{1}{n}p(w)\gamma \frac{w^2}{2}c(w)p(w)\\ & = & \frac{\gamma}{nw}\int_0^\infty dw \ ; [ p(w)]^2 c(w)\geq 0,\end{aligned}\ ] ] where we have integrated by parts .
the gini coefficient is thus proven to be a lyapunov functional of the fokker - planck equation .
it is well known that the equilibrium solution of the molecular boltzmann equation , namely the maxwell - boltzmann distribution , may be found by setting the variation of boltzmann s @xmath0 function to zero , under the constraints of fixed mass , momentum and energy .
we may attempt an analogous computation here , but , as conjectured by boghosian @xcite , the equilibrium solution of the boltzmann equation for the yard - sale model without redistribution is a singular generalized function , @xmath109 .
it is zero for @xmath110 , has zeroth moment @xmath38 and first moment @xmath39 , and its higher moments may not exist .
it is only when redistribution is included that steady - state solutions similar to the pareto distribution , eq .
( [ eq : parteopdf ] ) , are obtained @xcite .
still , it is instructive to see if the variational approach will yield this singular generalized function , so we turn our attention to that problem in this last section . using eq .
( [ eq : gini2 ] ) for the gini coefficient , @xmath49 $ ] , we introduce the lagrange multipliers @xmath111 and @xmath112 to enforce the constraints @xmath113 respectively .
we obtain @xmath114 - \lambda\int_0^\infty dw\ ; p(w ) - \mu\int_0^\infty dw\ ; p(w ) w\right ) \nonumber\\ & = & \frac{2}{w}\left[-w+\frac{1}{n}\int_0^w dx\ ; p(x)(w - x)\right ] - \lambda - \mu w , \label{eq : var}\end{aligned}\ ] ] where we have used eq . ( [ eq : frechet ] ) .
interpreting this in a weak sense , we take the zeroth moment of this with respect to @xmath12 . using eqs .
( [ eq : af ] ) , ( [ eq : gini1 ] ) and ( [ eq : gini2 ] ) , we obtain @xmath115 next , differentiating eq .
( [ eq : var ] ) yields @xmath116 - \mu \nonumber\\ & = & \frac{2}{w}\left[-1+f(w)\right ] - \mu \nonumber\\ & = & -\frac{2}{w}a(w ) - \mu , \label{eq : var1}\end{aligned}\ ] ] where we have used eq .
( [ eq : af ] ) . taking the first moment of this with respect to @xmath12 , and using eq .
( [ eq : gini2 ] ) , we obtain @xmath117 also , taking the limit as @xmath35 of eq .
( [ eq : var1 ] ) yields @xmath118 eqs .
( [ eq : mom0 ] ) , ( [ eq : mom1 ] ) and ( [ eq : momlim ] ) can be solved simultaneously to yield @xmath119 and , most importantly , @xmath68 .
so , under the dynamics of the yard - sale model , @xmath40 increases in time and asymptotically approaches the value one , corresponding to a state of `` perfect oligarchy . '' finally , note that differentiating eq .
( [ eq : var1 ] ) one more time immediately yields @xmath120 , valid everywhere expect @xmath33 , as expected .
we have proven that the gini coefficient @xmath40 is a lyapunov function of the boltzmann equation for the yard - sale model of asset exchange , as well as the fokker - planck equation obtained in the limit of small transaction sizes .
we have also shown that the equilibrium distribution , obtained in the time - asymptotic limit , is zero for all @xmath110 , and corresponds to @xmath68 .
as noted earlier , it is only when the yard - sale model is supplemented with a mechanism for redistribution that steady states similar to the pareto distribution are found . with redistribution , however , the gini coefficient is clearly no longer a lyapunov functional , since it would be possible to begin with a higher concentration of wealth than that obtained in the time - asymptotic limit
. it may be possible to find a different lyapunov functional for the boltzmann or fokker - planck equations with a redistribution term , but we leave this as a topic for future study . | in recent work @xcite , boltzmann and fokker - planck equations were derived for the `` yard - sale model '' of asset exchange . for the version of the model without redistribution ,
it was conjectured , based on numerical evidence , that the time - asymptotic state of the model was oligarchy
complete concentration of wealth by a single individual . in this work
, we prove that conjecture by demonstrating that the gini coefficient , a measure of inequality commonly used by economists , is an @xmath0 function of both the boltzmann and fokker - planck equations for the model . |
the probable success of experiments aimed at producing magnetically trapped ultracold molecular samples depends heavily on the effects of collisional processes .
for example , paramagnetic alkali dimers can be produced via photassociation of ultracold atoms @xcite , but the resulting molecules , typically in high - lying vibrational states , are subject to vibrational quenching collisions @xcite which can release a large amount of energy and dramatically affect the efficiency of the cooling .
alternatively , cold molecules in their vibrational ground states can be produced either by thermal contact with a cold helium buffer gas @xcite or by stark slowing , for species that possess an electric dipole moment @xcite .
collisions are of obvious importance to buffer - gas cooling ( bgc ) , as well as to forced evaporative cooling ( ec ) that will be required to lower the temperature of these gases further and achieve for instance bose - einstein condensation ( bec ) .
both processes require large elastic collision rates to thermalize the gas .
so far ec has not been realized in practice for molecules , but the success of the bgc technique for the production of cold cah @xcite and pbo @xcite molecules suggests that in the near future it will be possible to achieve bec using cold molecules .
this would open the way for a number of new and fascinating experiments . in order
to be magnetically trapped , atoms or molecules must be in a weak - field seeking state , i.e. a state whose energy increases with the strength of the magnetic field . for each trappable weak - field seeking state there
is in general a lower - energy untrapped strong - field seeking state , in which the molecules experience a force away from the center of the trap .
collisions can drive transitions between trapped and untrapped states .
these `` bad '' collisions can cause heating or atom loss .
it is therefore important to assess the rate constants for the inelastic collisions . in a series of previous papers
the resilience of molecular oxygen against spin - changing collisions was investigated , in collisions of o@xmath1 molecules both with a helium buffer gas @xcite , and with other o@xmath1 molecules @xcite .
these studies found that spin - changing rates due to spin - rotation coupling could be quite large .
however , in the case of the @xmath0o@xmath1 molecule , where in the limit of zero field the only allowed exit channels are energetically degenerate with the incident channels , spin - flipping transitions require boosting the centrifugal angular momentum from @xmath4 = 0 to @xmath4 = 2 , meaning that these processes are strongly suppressed by the wigner s threshold laws at collision energies smaller than the height of the exit channel centrifugal barrier . the results in refs .
@xcite considered only the case of a vanishing external magnetic field , which is obviously not the case in experiments that trap molecules using spatially inhomogeneous magnetic fields .
the present paper therefore explores the role that fields play in determining spin - changing collision rates .
as we demonstrate , the presence of a magnetic field causes a zeeman asymptotic splitting between incident and exit channels , thus lifting the collision energy higher relative to the centrifugal barrier in the exit channel , and removing the wigner s suppression .
studies of spin - changing ultracold collisions in the presence of an external magnetic field have been performed so far only for atomic species @xcite . in this paper
, we present a detailed dynamical study at cold and ultracold temperatures for the atom - diatom system @xmath0o@xmath1 @xmath5 @xmath2he in a field . the basic model is described in sec .
ii . in sec .
iii we calculate elastic and inelastic rate constants for collisions involving the lowest - lying trappable state of @xmath0o@xmath1 over a wide range of field values ( from 0 up to 5000 gauss ) , then discuss the dependence of the rates on collision energy for several representative values of the field .
this system is of direct relevance to bgc of molecular oxygen .
generally , it allows us to quantify the removal of the wigner s law suppression as the field increases in strength .
on this basis we determine a simple one - parameter fitting formula that reproduces the trend with field and energy of the loss rates . in sec .
iv we use this formula to extend previous results on o@xmath1-o@xmath1 collisions to estimate the influence of the field on ec of this system .
as mentioned in the introduction , we will consider in this paper molecules consisting of two @xmath0o atoms , whose nuclear spin @xmath6 is equal to @xmath7 .
we assume that total spin * i * = * i*@xmath8 + * i*@xmath1 is conserved in the collision and polarized to its maximum value * i * = 5 , implying that the even molecular rotational states @xmath9 are separated from the odd ones @xcite .
we limit the discussions of this paper to the `` even-@xmath9 '' manifold of molecular states , which is more appealing for cooling purposes @xcite having a spin 1 paramagnetic ground state .
the electronic spin * s * of the o@xmath1 molecule has magnitude @xmath10 = 1 in the electronic ground state @xmath11 we are concerned with throughout this paper .
the angular momentum * s * is coupled to the molecular rotation angular momentum * n * to give * j * , the total molecular angular momentum , which assumes the values @xmath12 , @xmath9 and @xmath13 for @xmath9 @xmath14 0 and is 1 for @xmath9 = 0 .
the hamiltonian operator @xmath15 for molecular oxygen in the presence of an external magnetic field @xmath16 can be written as : @xmath17 where @xmath18 is the rotational constant .
the @xmath0o@xmath1 molecule is considered to be a rigid rotor , with internuclear distance frozen to the equilibrium value of @xmath19 = 2.282 bohr ( the rigid rotor model has been shown to be very accurate for this system at the investigated collision energies @xcite ) .
the fine - structure hamiltonian @xmath20 and the hamiltonian @xmath21 for the interaction of the molecule with the external magnetic field follow the treatment in ref .
@xcite , disregarding the molecular hyperfine interaction .
the fine structure hamiltonian is given as @xcite @xmath22 where @xmath23 is a unit vector parallel to the molecular axis and @xmath24 is a second - rank tensor @xcite .
the fine structure parameters @xmath25 and @xmath26 have been taken from ref .
@xcite , where they have been determined by microwave spectroscopy .
the interaction of the field with the electronic spin can be expressed as @xcite : @xmath27 where @xmath28 indicates the external magnetic field , @xmath29 is the g - factor of the electron and @xmath30 is the bohr magneton .
following @xcite we ignore a small interaction between the field and the rotational angular momentum .
the matrix elements for @xmath20 and @xmath21 have been given in ref .
@xcite , eqs .
( a5 ) and ( a6 ) , for a hund s case b basis set .
we note here that the molecular rotational quantum number @xmath9 is no longer strictly a good quantum number for the molecular states , because different values of @xmath9 are coupled together by the fine - structure hamiltonian @xmath20 and by the interaction with the external field .
the molecular total angular momentum quantum number @xmath31 is still a good quantum number with respect to @xmath20 , but not with respect to the field interaction @xmath21 term .
however , its projection @xmath32 on the quantization axis is still conserved .
consequently , our basis functions should be labeled as @xmath33 , where @xmath34 is a shorthand index denoting the pair of quantum numbers ( @xmath35 ) in the field dressed basis @xcite . however , the coupling between different @xmath9 s is weak ( the fine - structure coupling is small compared to the rotational separation ) and @xmath9 can be considered `` almost '' a good quantum number .
similarly , @xmath31 is also approximately good for laboratory - strength magnetic fields , so that we can use without confusion the label @xmath36 .
magnetic trapping is strongly related to the behavior of the molecules in a magnetic field .
the low - energy zeeman levels of oxygen are displayed in fig . 1 for the even-@xmath9 species .
( throughout this paper we report energies in units of kelvin by dividing by the boltzmann constant @xmath37 .
these units are related to wave numbers via 1 k = 0.695 @xmath38 ) . in order to be trapped in the usual magnetic traps a molecule must be in a weak - field - seeking state , i.e. , one whose energy rises with increasing magnetic field strength .
thus the state @xmath39 = @xmath40 is the lowest - lying trappable state of the even-@xmath9 manifold , and this is the state on which we focus our attention below .
this state is indicated by a heavy line in figure 1 .
higher - lying states with @xmath9 @xmath41 2 are energetically forbidden at low temperatures .
it is clear from fig .
1 that for any trapped state there is an untrapped , strong - field - seeking zeeman state at a lower energy . these states are not merely untrapped but antitrapped , experiencing a force away from the trapping region . in a magnetic trap collisions with buffer gas atoms , or more generally with other molecules ,
will therefore ultimately deplete the trap of its molecular population .
the time available for cooling processes like bgc or ec , as well as the lifetime of a molecular bose - einstein condensate , is therefore limited and knowledge of the rate constants for spin - flipping collisions is essential to predict their feasibility . in ref .
@xcite the theoretical framework for atom - diatom scattering was derived in the limit of zero external field , along the lines of the model originally due to arthurs and dalgarno @xcite , and properly modified to incorporate the electronic spin of the oxygen molecule . here , the formulation of the scattering problem is further extended to account for the interaction with the external magnetic field .
the full hamiltonian operator describing the he - o@xmath1 collision is given by @xmath42 + { \bf \hat{h}}_{{\rm o}_2 } + v(r,\theta)\ ] ] after multiplying the wave function by @xmath43 in order to remove first derivatives . here
@xmath44 is the reduced mass for the he - o@xmath1 system , @xmath43 is the modulus of the jacobi vector joining the atom to the molecule s center - of - mass , @xmath45 is the centrifugal angular momentum operator , and @xmath15 is the molecular oxygen hamiltonian defined in ( [ o2ham ] ) . the potential term @xmath46 , depending on both the jacobi vector @xmath43 and on the bending angle @xmath47 that the molecule s axis makes with respect to @xmath48 , accounts for the he - o@xmath1 interaction .
we use the @xmath49 potential energy surface ( pes ) by cybulski et al .
@xcite , which approximates the true well depth to @xmath50 20% .
the full multichannel calculation requires casting @xmath51 in an appropriate angular momentum basis .
our field dressed basis for close - coupling calculations is then latexmath:[\[\label{basis } number @xmath10 is not explicitly indicated being always equal to 1 in the problem treated here . the quantum number @xmath4 stands for the partial wave representing the rotation of the molecule and the he atom about their common center of mass , @xmath53 is the projection of @xmath54 onto the laboratory axis , and @xmath55 is the laboratory projection of the total angular momentum , @xmath56 . at the collision energies of interest we assume that the oxygen electronic state and the helium atom state are preserved and
therefore we suppress the first two kets in equation ( [ basis ] ) in the following .
we note here that , at variance with the formulation in the zero field limit , the total angular momentum of the system @xmath57 ( equal to * n * + * s * + * l * ) is no longer a good quantum number , because different @xmath58 s are coupled together by the interaction with the external field .
this means that the dynamical problem is no longer factorizable for different values of @xmath58 , thus requiring larger numbers of channels to be treated simultaneously .
the problem is still factorizable for @xmath55 , but in general the number of channels to be included for each calculation is much larger than in the previous case .
numerical details of the calculations will be given in the next section .
the coupled - channel equations are then propagated using a log - derivative method @xcite and solved subject to scattering boundary conditions to yield a scattering matrix s : @xmath59 as already noted , the projection of the total angular momentum @xmath55 is still a good quantum number , implying that @xmath32 @xmath60 @xmath53 @xmath61 @xmath62 @xmath60 @xmath63 .
note that in general each of the quantum numbers @xmath9 , @xmath31 , @xmath32 , @xmath4 , and @xmath53 are subject to change in a collision , consistent with conserving @xmath55 .
following ref .
@xcite , the state - to - state cross sections are given by : @xmath64 and the corresponding rate coefficients are given by @xmath65 where @xmath66 is the relative velocity of the collision partners before the collision .
for notational convenience we will in the following refer to collisions that preserve the incident molecular quantum numbers as `` elastic , '' and those that change the quantum numbers as `` loss . ''
in this section we consider elastic and state - changing , inelastic rate constants for the incident channel @xmath67 = @xmath68 . at the investigated collision energies ( from 1 @xmath44k up to 10 k )
there are two open inelastic channels , namely @xmath67 = @xmath69 and @xmath70 , both of which are untrapped ( see fig .
we begin by computing rate constants for the elastic and spin - flipping transitions @xmath71
@xmath72 @xmath73 and @xmath71 @xmath72 @xmath74 in the low - field limit .
results in this section refer to 1 @xmath44k collision energy and are converged using partial waves up to @xmath4 = 6 and including the rotational states @xmath9 = 0 , 2 , 4 and 6 for the oxygen molecules .
the maximum value of @xmath43 to which the coupled - channel equations are propagated depends on the strength of the field , ranging from 600 bohr in the case of smallest field to 450 bohr for highest values .
these parameters assure rate constants convergent within less then 5 % , which is adequate for our purposes . in principle , the scattering matrix should be determined for each possible value of the projection of the total angular momentum @xmath55 .
however , we know that s - wave collisions dominate the incident channel at ultralow collision energies , which corresponds for our incident channel to the value @xmath75 . we have verified that including in the calculation only the @xmath55 = 1 channels changes the results by less than 1 % at @xmath44k energies ; in this section we therefore include only this contribution . the number of channels to be propagated according to the given convergent quantum numbers
is then only 205 .
2 shows elastic and inelastic rate constants at energy @xmath76k and for low values of the field .
the inelastic rate constants are non zero even at zero field , as shown in ref .
. however , in this limit the final states are degenerate with the initial state , and inelastic transitions are strongly suppressed by the presence of a @xmath77-wave centrifugal barrier in the exit channel , whose height is about 0.59 k. this effect is able to suppress the molecule loss , at least as long as the collision energy does not exceed the barrier height @xcite .
as soon as a field is applied , the thresholds are no longer degenerate in energy , so that the energy in the exit channel is not as far below the centrifugal barrier . as a consequence , inelastic transitions are no longer as strongly suppressed , even in the limit of very low collision energy .
rather , they increase dramatically even in a weak field , with rates being boosted by 5 or 6 orders of magnitude in a 1 gauss field . on the other hand ,
elastic scattering is nearly unchanged by the field .
this sudden increase of the inelastic transition rates can be reproduced semiquantitatively by applying the distorted wave born approximation ( dwba ) @xcite , as has been successfully done for the magnetic dipolar interaction of cold alkali atoms @xcite as well as in a number of problems in cold collisions @xcite .
the first order dwba is a simple two state perturbation approximation applicable in cases where inelastic scattering is weak in comparison with elastic scattering .
the dwba expression for the off - diagonal k - matrix elements is : @xmath78 where @xmath79 and @xmath80 represent the energy normalized scattering wave function of the initial and final states calculated on the diabatic potential corresponding to the states involved in the inelastic transition .
@xmath81 is the corresponding diabatic coupling term of the hamiltonian ( [ totalh ] ) .
the first - order dwba result is shown as a dotted line in fig.2 .
the dwba also yields information on the threshold dependence of the loss rates on energy and field . to see this
, first note that the spin - changing processes we are considering are strongly dominated at low collision energy by s - waves in the incident channel , and by d - waves in the exit channel .
this change of partial wave is necessary to conserve angular momentum during a collision that changes the molecule s spin . for small values of the magnetic field
( for which the zeeman splitting does not exceed the height of the exit channel centrifugal barrier ) the exit channel is still in the threshold regime , whereby the wave functions @xmath79 and @xmath80 in ( [ kmatr ] ) can be approximated by the small - argument limit of energy - normalized spherical bessel functions , @xmath82 where @xmath83 and @xmath84 are the incident and final wave numbers and @xmath85 and @xmath86 are the incident and final partial waves .
assuming a small-@xmath43 cutoff to insure convergence of the integral in ( [ kmatr ] ) with respect to @xmath87 singularity in the coupling potential @xmath88 , the energy - dependence of the k - matrix element is @xmath89 by considering the relationship between @xmath90 and the effective rate constant @xmath91 , it is straightforward to show that the rate constant behaves approximately as @xmath92 where @xmath93 is the collision energy , and we have taken into account that the final kinetic energy in the exit channel is incremented by an amount @xmath94 = @xmath95 @xmath29 @xmath96 corresponding to the linear zeeman shift .
@xmath95 (= @xmath32 @xmath5 @xmath62 ) stands for the difference between the initial and final values of @xmath32 in the two channels involved in the transition .
the actual value of @xmath94 is modified by quadratic zeeman shift , but the linear approximation is adequate for achieving a simple fitting formula . in the present case
these shifts amount to less than 10% changes in the approximated rate constants . from eq .
( [ prope ] ) , considering that in our case @xmath97 = 0 and @xmath98 = 2 , a simple expression for the rate constants can be derived : @xmath99 where @xmath100 represents an overall scaling constant and @xmath101 is conveniently chosen as the height of the centrifugal barrier in the exit , d - wave channel . in the limit of very low collision energy ,
the @xmath102 term obviously dominates over @xmath93 , leading to a nonzero rate constant , as is the case for exothermic collisions .
this simple expression allows us to interpret the threshold behavior of the rates with the field , explaining the 5/2 exponential dependence on @xmath16 found in our calculation and shown explicitly in the bilogarithmic plot for the rate constants ( fig .
the dashed lines represent a fit to the rate constants in the limit of zero magnetic field , yielding coefficients @xmath100 = 2.73 @xmath103 10@xmath104 cm@xmath2 sec@xmath106 and @xmath100 = 1.45 @xmath103 10@xmath104 cm@xmath2 sec@xmath106 for the transition to the final states @xmath73 and @xmath74 , respectively .
apart from zeros in the actual rate constants , the overall trend is indeed @xmath107 .
the zeros in the real rates arise from interferences between the s - and d - wave radial wave functions , as we have verified qualitatively by the dwba .
nevertheless , the simple one - parameter expression ( [ ratek ] ) provides a reasonable upper bound to the complete calculation which , it will be recalled , requires a calculation involving 205 coupled channels .
the simple formula ( [ ratek ] ) holds , of course , only when both incident and final channels are in the threshold regime . assuming low incident energies
, this restriction therefore limits the size of magnetic field for which ( [ ratek ] ) applies .
namely , this expression is only useful when the zeeman energy splitting between incident and final states remains smaller than the height of the d - wave centrifugal barrier . for the channels considered here , these fields are 2430 gauss and 4860 gauss for the @xmath108 and @xmath109 final states , respectively ; vertical arrows in fig .
3 indicate these field values .
relation ( [ ratek ] ) serves as a useful quick fitting formula for data at smaller fields , and allows us to generalize the results obtained in this paper also to different systems .
possible applications are discussed in section iv . when stronger values of the magnetic field are considered , the dwba is no longer able to reproduce the full calculation , because the coupling is no longer a weak perturbation .
strong field interaction mixes up different channels leading to a much more complicated picture which can not be explained in terms of a simple two state model .
however , in this limit the loss rate constants seem to be unacceptably large anyway , owing to the effectiveness of spin - rotation coupling in changing spins .
for example , for @xmath16 @xmath50 4800 gauss , the inelastic rate constant for the @xmath67 = @xmath73 exit channel exceeds that for the elastic one .
this indicates that the high field limit fine structure changing collisions represent a potential limitation for the success of the collisional cooling processes . in this section ,
the dependence of elastic and inelastic rate constants on the collision energy are analyzed for the zero field limit and for three representative values of the field , namely 10 , 200 and 4500 gauss . in fig .
4.a , we report our results for collision energies 1@xmath44k - 1 k , for the four mentioned values of the field
. elastic scattering , which is largely determined by s - waves in both incident and final channels , is weakly affected by the strength of the field . for spin - changing collisions , however , the energy dependence changes dramatically , in accordance with ( [ ratek ] ) , which is shown in the figure using dashed lines . again
the trends are well - represented , although the formula overestimates the rates due to the zeros in the real rates described in the previous section . as a general trend
, we observe that when the field is increased , the low - energy inelastic rate constants are substantially pushed up towards the elastic ones , but at higher energies the rates are less sensitive to the field .
this is better illustrated in fig .
4.b , where the total loss rate constants ( that is the sum of the inelastic scattering in the two exothermic channels ) for the different values of the field are plotted together with the elastic channel results in zero field ( which , as stressed before , are essentially independent of the @xmath16 value ) . in the range of collision energies from 1@xmath44k to 1 k and for low magnetic fields
the rates for elastic collisions remain significantly higher than the rate for the spin - flipping . in particular , at buffer - gas - cooling energies of @xmath110k and below , the inelastic rate constants are in the @xmath111 cm@xmath2/sec range , and remain so even in the presence of a field .
this result verifies the suitability of the @xmath0o@xmath1 molecule for bgc . given the comparatively small uncertainties in the pes @xcite , the results shown here are probably fairly realistic for the he - o@xmath1 system .
we have continued the analysis in the range of collision energies from 1 to 10 k. we note that for energies larger than the height of the centrifugal barrier , the approximation of including only @xmath55 = 1 in the calculation is no longer accurate within a few percent , as was the case at lower energy .
we checked that in this energy range the full calculations including all the possible @xmath55 values provides results which differ ( in the worst case ) by about a factor of 3 for the inelastic channels and by about a factor of 5 for the elastic one .
the full calculation is computationally very expensive for @xmath16 @xmath14 0 , as opposed to the case of zero field where a total-@xmath58 basis can be adopted .
results indicate that inelastic rates for high collision energies are not much lower than elastic rates , and of course become higher still at energies where resonances exist . at collision energies above 1k
both feshbach and shape resonances appear in the cross sections , as noted in @xcite .
we find that these resonances move somewhat as a function of magnetic field .
nevertheless , they are sufficiently narrow that they are completely `` washed out '' by thermal averaging in a gas .
we therefore present these results as a function of temperature rather than energy . to this end
we assume a maxwellian velocity distribution of the collision partners characterized by a kinetic temperature @xmath112 .
the thermally averaged rate constants are then expressed as : @xmath113 where @xmath37 is the boltzmann constant and @xmath114 stands for cross sections . to compute this average ,
the values of the cross sections for @xmath93 @xmath14 10 k are extrapolated from their values at 10 k. averaged rate constants are shown in fig . 5 for the same set of field values as in fig .
4 . the condition for magnetic trapping to be successful is usually expressed as @xmath115 @xmath14 10 @xmath116 @xcite , so that we can conclude that for collisions of @xmath0o@xmath1 with @xmath2he this condition is fulfilled at least for temperatures up to 1 kelvin and values of the field for which the asymptotic zeeman splitting does not exceeds the height of the exit channel centrifugal barrier .
even in the absence of detailed information on cold collisions of a particular molecule , the fitting formula ( [ ratek ] ) can be used as an approximate guide to what the rates might be .
for example , we can inquire about the prospects for evaporatively cooling @xmath0o@xmath1 molecules once they have been successfully cooled in a first stage of bgc .
for this system , the d - wave centrifugal barrier height is @xmath117 mk , meaning that the threshold law is expected to hold for field values smaller than @xmath50 53 gauss for transitions that produce one or more molecules in the @xmath73 final state . in the case of @xmath0o@xmath1 cold collisions
we have access to the zero - field calculations of ref .
we have fit the energy dependence of all the inelastic rate constants to yield the dashed curve labeled `` @xmath118 '' in figure 6 , which represents the total loss .
the full calculation ( solid line ) has some additional features due to scattering resonances near zero energy , but this will not affect our conclusions here . based on the zero - field fit , we use ( [ ratek ] ) to estimate the loss rates in nonzero field .
the general trends are the same , namely , the rates rise sharply at low energy , but are roughly field - independent at larger energies . for evaporative cooling to be successful requires , roughly speaking , that the ratio of elastic to inelastic collision rates , @xmath119 , exceed 100 @xcite . for the estimated results shown in figure 6
, this condition holds only at energies below @xmath110mk for fields as low as 10 gauss , and not at all for near - critical fields of @xmath120 gauss . thus evaporative cooling from an initially buffer - gas - cooled sample may prove trickier than previously expected .
on the other hand , bgc is characterized by a large number of molecules cooled in the initial step ; it is possible that a certain loss can be tolerated , and that a final sample at sub-@xmath44k temperatures will still hold enough molecules to reach critical phase space density for bec .
detailed kinetic simulations are required to determine if this is so .
alternatively , a recent proposal suggests that nh molecules could be cooled via stark slowing to temperatures as low as 1 mk @xcite . in this case
ec may work quite well .
for many systems of interest to ultracold studies , there does not exist any information on spin - changing collisions at low enough temperatures .
in such cases it may be possible to make order - of - magnitude guesses anyway .
for example , suppose a rate constant is known for higher temperature collisions . in the absence of any other information we could simply assert that the rate has the same value at the energy @xmath101 corresponding to the height of the centrifugal barrier .
the fitting formula ( [ ratek ] ) then gives the behavior of this rate at lower temperatures .
for example , the zero - field rate constant in the first panel of fig .
5 has a value of @xmath121 cm@xmath2/sec at a temperature of 4 k. extrapolating this value down to @xmath122 k yields a coefficient @xmath123 cm@xmath2/sec for the total loss , just a factor of two higher than the fit to the low - energy calculation .
one of the main aims of this paper is to understand in a broad sense collision of paramagnetic molecules in the limit of ultralow temperatures in the presence of a magnetic field .
this is a part of our effort in showing that molecules with nonzero spin in the lowest energy state ( such as @xmath0o@xmath1 investigated here ) can be successfully cooled and used for bec purposes .
elastic and inelastic scattering in presence of a magnetic field for the specific system @xmath2he - @xmath0o@xmath1 has been characterized in detail .
our attention has been focused on the lowest lying trappable state of the molecule .
this information is immediately relevant for bgc of molecular oxygen , and suggests that the presence of the field does not particularly hinder the bgc process .
this work extends previous predictions which referred to the oxygen molecule in zero field @xcite , and definitely assess the theoretical possibility for trapping this species .
moreover , we have illustrated the simple underlying physics of spin - changing rates in general .
for fields low enough that both the incident and exit channels are in the threshold regime , the rate constants vary according to the wigner - law dependence ( @xmath93 + @xmath102)@xmath124 .
this insight allows us to make estimates of rate constants beyond the ones calculated in detail .
in particular , a small field was found to have a dramatic effect on the evaporative cooling of @xmath0o@xmath1 , which must be taken into account in future experiments aimed at quantum degenerate molecular gases . | we investigate the collisional stability of magnetically trapped ultracold molecules , taking into account the influence of magnetic fields .
we compute elastic and spin - state - changing inelastic rate constants for collisions of the prototype molecule @xmath0o@xmath1 with a @xmath2he buffer gas as a function of the magnetic field and the translational collision energy .
we find that spin - state - changing collisions are suppressed by wigner s threshold laws as long as the asymptotic zeeman splitting between incident and final states does not exceed the height of the centrifugal barrier in the exit channel .
in addition , we propose a useful one - parameter fitting formula that describes the threshold behavior of the inelastic rates as a function of the field and collision energy .
results show a semi - quantitative agreement of this formula with the full quantum calculations , and suggest useful applications also to different systems . as an example , we predict the low - energy rate constants relevant to evaporative cooling of molecular oxygen . # 1@xmath3#1 |
increasing disorder is known to suppress superconductivity in low - dimensional systems such as thin films and narrow wires@xcite .
this occurs because the disorder causes electrons to move diffusively rather than ballistically , making them less efficient at screening the coulomb repulsion between electrons .
the increased coulomb repulsion decreases both the electron - electron attraction needed for superconductivity , and the density of states of electrons available for pairing at the fermi surface@xcite .
typical types of experimental data are : ( i ) @xmath4 , the transition temperature as a function of normal state resistance per square@xcite ; ( ii ) @xmath5 , the order parameter at zero temperature , as a function of normal state resistance per square@xcite ; ( iii ) @xmath6 , the upper critical field as a function of temperature and normal state resistance per square@xcite ; ( iv ) @xmath7 , transition temperature as a function of resistance per square and spin - flip scattering rate in films with magnetic impurities.@xcite it is found experimentally that @xmath4 curves from a wide variety of materials fit a universal curve with a single fitting parameter , whilst the few experimental measurements of @xmath8 , seem to have @xmath9 roughly constant@xcite .
this fitting to a single curve , whilst pleasing in showing that the basic ingredients of our theories are correct , does not allow detailed analysis of the theory .
data of types ( iii ) and ( iv ) are more promising because there is an additional parameter to vary the magnetic field in ( iii ) , and spin - flip scattering rate in ( iv ) . to the best of our knowledge ,
only one experiment of type ( iv ) has been performed@xcite , and we discuss it elsewhere.@xcite several experiments of type ( iii ) have been performed@xcite ; some seem to show a positive curvature in @xmath0 at low temperature as disorder is increased .
moreover this effect is predicted by theory@xcite , and this seems to be another confirmation of the basic theoretical model .
however , we need to be careful : positive curvature in @xmath0 is a ubiquitous feature of exotic superconductors@xcite , and occurs in many systems where localization is not believed to be the cause .
indeed any pair - breaking mechanism that varies as a function of magnetic field can lead to such anomalous behaviour in @xmath2 .
this means that it is often difficult to distinguish between the various mechanisms that might be present .
it is therefore particularly important to be sure of our theory , and in this light we re - examine the predictions of localization theory .
one of the main problems of the localization theory is that even first - order perturbation theory results are hard to obtain correctly .
the first - order results are capable of explaining experimental data in the weak disorder regime , but for stronger disorder it is clear that we need something else .
as an example consider the prediction for @xmath10 suppression@xcite , @xmath11 where @xmath12 is transition temperature for clean system , @xmath13 , and @xmath14 is the elastic scattering time .
this yields an exponential curve for @xmath4 , which behaves like a straight line for small @xmath1 .
it is clear that @xmath10 deduced from this equation can never go to zero for finite @xmath1 , as happens in experiment .
a very simple ad hoc way of going beyond simple perturbation theory is to replace @xmath12 on the right hand side by @xmath10 , pleading perhaps to self - consistency .
if we define @xmath15 , @xmath16 , and @xmath17 , the new equation has the cubic form @xmath18 and can easily be solved .
however a new problem emerges because there are two positive roots for every value of @xmath1 . at first
we can take the larger of the roots , on physical principles , because it is this root which tends to @xmath19 at @xmath20 .
however we eventually come to a re - entrance point beyond which no solutions exist .
it is clear that this re - entrance is unphysical , an artefact of our ad hoc extension of perturbation theory . in the case of @xmath4
the story has a happy ending in that perturbation theory can be correctly extended by a renormalization group ( rg ) treatment based on finkelstein s interacting non - linear sigma model@xcite .
this leads to the result @xmath21 where @xmath22 .
this equation reduces to the first - order result for small @xmath23 and now @xmath10 goes smoothly to zero at @xmath24 .
the three curves are plotted for comparison in fig .
the reason for discussing @xmath4 in detail above is that the same problem occurs for @xmath6 .
the standard theory in this case@xcite , due to maekawa , ebisawa and fukuyama ( mef ) is the equivalent of the ad hoc extension discussed above , and has the form @xmath25\right)-r_{hf}-r_v
\nonumber\\ r_{hf}&=&\displaystyle-{1\over 2}t\ln^2{\left({1\over 2\pi t_c\tau}\right ) } -t\ln{\left({1\over 2\pi t_c\tau}\right ) } \left[\psi\left({1\over 2}\right )
-\psi\left({1\over 2}+{1\over 2\pi t_c\tau_h}\right)\right]\nonumber\\ r_v&=&\displaystyle-{1\over 3}t\ln^3{\left({1\over 2\pi t_c\tau}\right ) } -t\ln^2{\left({1\over 2\pi t_c\tau}\right ) } \left[\psi\left({1\over 2}\right ) -\psi\left({1\over
2}+{1\over 2\pi t_c\tau_h}\right)\right],\end{aligned}\ ] ] where @xmath26 .
this equation suffers similar re - entrance problems at finite @xmath27 to those found at @xmath28 , when it is just the @xmath10 equation . indeed
the @xmath0 curves can only be plotted down to the value of @xmath3 at which re - entrance occurs , and at this point the curves appear to have infinite slope .
this leads to us asking the question of whether the positive curvature in @xmath0 is also an artefact of the ad hoc approximation used .
what we need to answer this question is the finite magnetic field analogue of the rg result discussed above .
however the rg is very difficult , and the answer is not forthcoming from this source .
fortunately oreg and finkelstein@xcite have recently shown that the rg result can be obtained from a diagrammatic resummation technique .
this method has the great advantage of being analytically tractable and easy to use . in this paper
we extend this approach to finite magnetic field to see whether we really do expect positive curvature in @xmath0 .
indeed this paper is intended as somewhat of a showcase for this resummation method , to demonstrate the ease of its extension to a wide variety of problems .
oreg and finkelstein@xcite take the coulomb interaction to be featureless , and we shall follow them . we should therefore explain why it is legitimate to use a featureless interaction rather than the correct screened coulomb interaction .
the screened coulomb interaction has a singularity at low momentum , and one might naively think that this would lead to a strong enhancement of the suppression of transition temperature .
however a cancellation occurs between diagrams 14 and diagram 5 of fig .
2 , which effectively removes the low momentum singularity@xcite .
the coulomb interaction is then effectively featureless , all diagrammatic sums being dominated by large frequency and momentum .
it turns out that one gets the same result from doing the perturbation theory correctly with the screened coulomb interaction as from using a constant interaction of strength @xmath29 and keeping only diagrams 14 .
this is what we shall do in this paper .
we also note that the re - summation method of oreg and finkelstein@xcite keeps only diagrams 3 and 4 , which is again legitimate as their contribution is greater than that from diagrams 1 and 2 .
however it is not difficult to include these terms in the re - summation , as we will show in section .
the cancellation of low momentum singularities is expected to be a general feature , enforced by gauge invariance , so we are free to ignore them and use a featureless interaction as long as we include the diagrams that give the dominant contribution at large frequency and momentum .
the outline of the rest of the paper is as follows . in section
we review the first - order perturbation calculation for @xmath0 assuming a featureless coulomb interaction .
we derive the analytical mef formula by making an asymptotic approximation to the matsubara frequency sums which arise in perturbation theory .
we make plots of @xmath0 both by solving the implicit mef equation , and its equivalent where the matsubara sums are performed exactly .
surprisingly we find no positive curvature in the latter case .
in section we derive @xmath0 using the resummation technique both in the form used by oreg and finkelstein , and also in an extended form which includes self - energy diagrams missed in their formalism .
again we find no positive curvature in @xmath0 . in section
we discuss the experimental situation and draw conclusions .
in this section we will carefully review the first - order perturbation theory calculation@xcite of @xmath0 .
we do this in detail because we will show that accurately performing sums rather than making an analytical approximation to them removes the upward curvature in @xmath0 .
we identify @xmath10 as the temperature at which the magnetic field dependent pair propagator diverges . to find the latter we calculate the correction to the momentum - dependent pair propagator , @xmath30 , and then make the usual substitution @xmath31 , where @xmath26 . the zeroth - order ( mean field )
pair propagator is given by @xmath32\ ] ] so that upon substitution @xmath31 we get the usual abrikosov - gorkov result@xcite for the pair - breaking effect of the magnetic field @xmath33 we will calculate the corrections to the pair polarization bubble , @xmath34 , which will lead to a change in @xmath10 given by @xmath35 since we will assume a featureless interaction , we need to evaluate diagrams 1 to 4 of fig .
diagrams 1 and 3 involve the summation of three terms to form a hikami box@xcite , the general form for which is @xmath36 where @xmath37 and @xmath38 , the @xmath39 being the incoming momenta , and the @xmath40 the matsubara frequencies on the electron green functions in the box . using standard diagrammatic rules , and the above result for the hikami box , we obtain the 4 contributions to the pair polarization bubble @xmath41 we note that the relative signs of @xmath42 and @xmath43 are irrelevant in the sum , since the summand depends only upon @xmath44 and @xmath45 , and that the featurelessness of the potential allows us to shift the momentum sum in terms 2 and 4 . we find that terms 1 and 2 partially cancel , whilst terms 3 and 4 reinforce , to yield the sum @xmath46[\epsilon_l+\epsilon_n]}+ { 1\over[\epsilon_l+1/\tau_h][\epsilon_n+1/\tau_h ] } \left[\ln{\left({1\over[\epsilon_l+\epsilon_n]\tau}\right ) } + { 1\over[\epsilon_l+\epsilon_n]\tau_h}\right]\right\}\ ] ] where @xmath47 and we have performed the @xmath48-sum and set @xmath31 . we first reproduce mef s analytic approximation to the sums over matsubara frequencies . to do this it turns out to be easier to make a choice of relative sign of matsubara frequencies , @xmath49 , and to set @xmath50 . if @xmath51 and @xmath52 , the sum becomes @xmath53m}+ { 1\over[l+1/2+\alpha][m - l-1/2+\alpha ] } \left[\ln{\left({m\over m}\right)}+{\alpha\over m}\right ] \right\}\nonumber\\ & = & \displaystyle-{g\over 4\pi^2d}\sum_{m=1}^{m } \left[\psi\left({1\over 2}+m+\alpha\right ) -\psi\left({1\over 2}+\alpha\right)\right ] \left\{{1\over m}+{2\over ( m+2\alpha)}\left[\ln{\left({m\over m}\right ) } + { \alpha\over m}\right]\right\}\end{aligned}\ ] ] where @xmath54 , @xmath55 .
if we first evaluate the sum for @xmath56 , we see that it will be dominated by large @xmath57 , so we replace the difference of digamma functions by @xmath58 , and the sum over @xmath57 by an integral to get @xmath59.\ ] ] to evaluate the result for finite @xmath60 we subtract off the @xmath56 result and try to analytically approximate the difference .
again we expect the sum to be dominated by large @xmath57 , so we can ignore the @xmath60 in @xmath61 and @xmath62 , and ignore the term proportional to @xmath60 as it is less divergent .
the difference then has the form @xmath63
\left\{{1\over m}+ { 2\over m}\ln{\left({m\over m}\right)}\right\}\nonumber\\ & = & \displaystyle-{g\over 4\pi^2d } \left[\psi\left({1\over 2}\right)-\psi\left({1\over 2}+\alpha\right)\right ] \left\{\ln{\left({1\over 2\pi t_c\tau}\right)}+ \ln^2{\left({1\over 2\pi t_c\tau}\right)}\right\}\end{aligned}\ ] ] from our knowledge of the calculation which includes the full screened coulomb interaction we know that we should set @xmath64 , from which it follows that @xmath65 .
putting the eqns .
( [ mefpapprox ] ) and ( [ mfpapprox ] ) into eqn .
( [ tch ] ) yields the mef formula cited in eqn .
( [ mef ] ) .
it turns out that this approximation , although apparently reasonable , is not justified .
if we plot the values of @xmath66 calculated by performing the sums directly , we find that they do not agree with eqn .
( [ mef ] ) .
consequently the @xmath0 curves predicted by the analytic approximation and exact sum are also different . 0.25truein on the left side of fig .
3 we display @xmath0 predicted by the mef result of eqn .
( [ mef ] ) for several values of @xmath1 in the case where @xmath67 .
we clearly see the positive curvature at low temperature , and the result that the curves all terminate at finite @xmath3 due to re - entrance problems . on the right side of fig .
3 we plot @xmath0 deduced directly from the first - order perturbation theory result of eqn .
( [ passum ] ) .
we note that since the answers we get depend upon the logarithm of the upper cut - off , if we treat the upper cut - off differently , we will get slightly different answers .
however , this difference will always be lost in fitting to experiment since we determine the unknown upper cut - off parameter from the initial slope of the data .
the reason we mention this is that the exact sums can be performed in two slightly different ways : we can have the sum over @xmath42 and @xmath68 , or the sum over @xmath42 and @xmath43 , the cut - offs in each case being taken at @xmath69 .
the results are slightly different due to different treatment of the upper cut - off . however in both cases
we find no positive curvature in @xmath0 .
there is still the problem that the curves finish at finite @xmath3 because of re - entrance , and this tells us that the ad hoc prescription being used to go beyond first - order perturbation theory is still inadequate .
but again we stress the key point of this section : even within the ad hoc extension of first - order perturbation theory , we do not get positive curvature in @xmath0 if we do the matsubara sums exactly .
in this section we answer the question of what is the correct way to go beyond first - order perturbation theory . to do this , we extend a resummation technique recently developed by oreg and finkelstein@xcite ( of ) .
this involves calculating the pair scattering amplitude , @xmath70 , and identifying @xmath10 as the temperature at which it first diverges .
the ladder summation involved is shown diagrammatically in fig .
4 this is actually an extended version of the of approach because as well as considering diagrams 3 and 4 of fig . 2 , which correspond to an effective coulomb pseudopotential , and end up in the block @xmath71 , we also consider diagrams 1 and 2 of fig . 2 , which correspond to an effective self - energy , and end up in the block @xmath72 which renormalises the cooperon impurity ladder .
in fact we will consider both versions of the summation technique : the simpler one involves using the bare cooperon @xmath73 rather than the dressed one @xmath74 .
of demonstrate that the matrix equation they obtain in 2d can be approximated by a parquet - like differential equation that turns out to be identical to that obtained from renormalization group analysis . instead of generating a continuous approximation to the matrix equation , we simply solve the equation and change the temperature ( which affects the upper cut - off @xmath55 ) until we find the first zero eigenvalue of the matrix , at which point it has become singular , and we have reached @xmath10 . numerically this involves diagonalizing matrices of rank less than 2000 or so .
let us now proceed to the calculation of the pair amplitude matrix @xmath75 . if we evaluate the diagrams of fig . 4 ,
but ignore the magnetic field and self - energy correction to the cooperon , we obtain the same equation as of , @xmath76 { 1\over |\epsilon_m|}\gamma(\epsilon_m,\epsilon_l)\ ] ] except that we have explicitly kept both positive and negative matsubara frequencies because the expression for @xmath77 will turn out to depend upon whether its two matsubara frequencies have the same or opposite sign .
this is a reflection of the breaking of time - reversal invariance caused by the magnetic field .
let us first see how to solve the above equation , and then later put in the magnetic field , and the correction to the cooperon . as a matrix equation , it takes the form @xmath78{\hat\epsilon}^{-1}{\hat\gamma}\ ] ] where @xmath79 , @xmath80 , @xmath81 and @xmath82 .
this has the solution @xmath83 where @xmath84{\hat\epsilon}^{-1/2}\ ] ] and @xmath85 is the identity matrix .
it follows that when the matrix @xmath86 has an eigenvalue equal to @xmath87 , the pair amplitude diverges , and we have found @xmath10 .
our approach is therefore to start at the mean field transition temperature , @xmath12 , and decrease temperature until one of the eigenvalues of @xmath88 changes sign .
the matrix @xmath89 depends upon @xmath3 both through the dependence of @xmath90 upon @xmath3 and through its rank @xmath91 , where @xmath92 .
we start at the mean - field value of @xmath93 , which we will call @xmath94 , and decrease the temperature by increasing @xmath93 successively by one .
we diagonalize the matrix @xmath95 for each value of @xmath93 until an eigenvalue changes sign . at this point
we have found @xmath10 for the given problem , and @xmath96 . we can then change a parameter such as @xmath23 or @xmath60 and repeat the procedure .
this removes the need for any perturbative expansion .
let us now put back in the magnetic field and cooperon self - energy corrections .
the @xmath97 denominator in the eqn .
( [ of ] ) comes from the cooperon @xmath98 and in the presence of a magnetic field , @xmath97 is replaced by @xmath99 , and hence @xmath100 , where @xmath101 .
the contributions to @xmath71 from the diagrams of fig . 4 which correspond to diagrams 3 and 4 of fig .
2 are given by @xmath102}\theta(-\epsilon_n\epsilon_m)\nonumber\\ t\lambda_4(\epsilon_n,\epsilon_m)&=&\displaystyle { g\over\pi n(0)}\sum_{q'}{1\over [ dq'^2+|\epsilon_n|+|\epsilon_m| ] } \theta(\epsilon_n\epsilon_m)\end{aligned}\ ] ] and so performing the @xmath48-sum and setting @xmath31 gives @xmath103 } & \quad$\epsilon_n\epsilon_m>0$\cr \displaystyle\ln{\left[{1\over ( |\epsilon_n|+|\epsilon_m|)\tau}\right ] } + { 2\over ( |\epsilon_n|+|\epsilon_m|)\tau_h}&\quad $ \epsilon_n\epsilon_m<0$\cr}\ ] ] it follows that the matrix elements of @xmath90 are @xmath104 to include the self - energy correction into the cooperon , we note that the corrected cooperon is given by @xmath105^{-1}= { 1\over 2\pi n(0)\tau^2}\displaystyle \left[dq^2 + 2|\epsilon_m|-{1\over 2\pi n(0)\tau^2}\sigma\right]^{-1}\ ] ] so that we need to absorb a factor @xmath106 into @xmath72 .
the contributions from the diagrams of fig . 4 which correspond to diagrams 1 and 2 of fig . 2
are @xmath107\theta(-\epsilon_n\epsilon_m ) \nonumber\\ \sigma_2(\epsilon_n)&=&\displaystyle-{2g\over n(0)}t\sum_{\epsilon_m}\sum_{q ' } { 1\over ( dq'^2+|\epsilon_n|+|\epsilon_m|)}\theta(\epsilon_n\epsilon_m)\end{aligned}\ ] ] which partially cancel to give the result @xmath108\ ] ] the weak - localization contribution to the cooperon self - energy is given by @xmath109\ ] ] incorporating @xmath72 into the cooperon means that @xmath110 has elements @xmath111 -2\alpha t\ln{\left({m\over n+1/2}\right)}\right\}\delta_{nm}\ ] ] using the new formulas for @xmath110 and @xmath90 , we now plot @xmath0 for various values of @xmath17 .
the results are shown in fig . 5 :
the plot on the left does not include self - energy corrections ; the plot on the right does .
the results are very similar and show no sign of upward curvature in @xmath0 .
we note that there are no re - entrance problems in our calculation of @xmath10 : we can plot the curves down to as low a temperature as we like if we are prepared to diagonalize large enough matrices . 0.25truein
the central message of the paper is that the theory of localization and interaction does not predict an anomalous positive curvature in the upper critical field @xmath0 of thin film superconductors .
a subsidiary message is that the resummation method developed by oreg and finkelstein@xcite is a very powerful and adaptable tool for going beyond perturbation theory in disordered superconductors .
we suspected that the positive curvature in @xmath0 was an artefact related to the re - entrance problem found in @xmath4 , and that it would not survive a systematic non - perturbative treatment .
the latter is true , but the cause of the positive curvature artefact turns out to be an incorrect analytic approximation to sums of matsubara frequencies .
we will now consider the experimental situation by attempting to fit experimental data to both the mef formula of eqn .
( [ mef ] ) , and the exact first - order perturbation theory result of eqn .
( [ passum ] ) .
the data which fits best to the localization theory is that of graybeal and beasley@xcite on amorphous films of mo - ge , because values of @xmath10 for all films can be obtained by choosing a single value of the parameter @xmath16 .
the fits in fig .
6 used the parameter value @xmath112 for the mef curves and @xmath113 for the exact sum curves .
values of @xmath114 and @xmath115koe were taken directly from the data .
the experimental data appears to fit better to the exact sum curves than the mef curves , and there seems little sign of an upward curvature in @xmath0 .
when the same procedure is applied to the data of okuma et al@xcite on zn films , we find that we need to use different values of @xmath116 for the two films to get the correct @xmath10 for given @xmath1 .
the plots in fig . 7 use values @xmath117 for the @xmath118 film and @xmath119 for the @xmath120 film for the mef formula ; @xmath121 for the @xmath118 film and @xmath122 for the @xmath120 film for the exact sum formula .
both films have the same thickness of @xmath123 , but different resistances , and hence different diffusion constants @xmath124 .
the latter were calculated using the material parameters for zn to get values of @xmath125 for zero - resistance films of the same composition equal to 5.08koe for the @xmath118 film and 7.62koe for @xmath120 film .
@xmath12 was taken to be @xmath126 .
again the data seems to fit better to the exact sum curves than the mef formula .
to fit the data of hebard and paalanen@xcite on amorphous in - ino@xmath127 films again requires a different value of @xmath116 for each film .
the plots in fig .
8 use @xmath128 for the @xmath129 film and @xmath130 for the @xmath131 and @xmath132 films for the mef curves ; @xmath133 for @xmath129 film and @xmath134 for the @xmath131 and @xmath132 films for the exact sum curves .
the value of @xmath135 quoted in the paper was used , and a value of @xmath136 was estimated from the experimental plots .
there is a clear upturn in @xmath0 at low @xmath3 , but it is not explained by the mef curves which are effectively all upturn at these parameter values . again
the fit to exact sum curves seems better . in conclusion
it seems that the experimental curves fit better to the exact sum curves than the mef curves .
any low temperature upturns in @xmath0 are not explained by the mef formula , and presumably correspond to different physics .
we thank i. aleiner , a. clerk , a.m. finkelstein , i.v .
lerner and y. oreg for helpful discussions .
ras acknowledges the support of a nuffield foundation award to newly appointed lecturers in science and mathematics .
bsh acknowledges the support of a uk epsrc graduate studentship .
va is supported by the us national science foundation under grant dmr-9805613 . | we present calculations of the upper critical field @xmath0 in superconducting films as a function of increasing disorder ( as measured by the normal state resistance per square @xmath1 ) . in contradiction to previous work , we find that there is no anomalous low temperature positive curvature in @xmath2 vs @xmath3 as disorder is increased .
we show that the previous prediction of this effect is due to an unjustified analytical approximation of sums occuring in the perturbative calculation .
our treatment includes both a careful analysis of first - order perturbation theory , and a non - perturbative resummation technique .
no anomalous curvature is found in either case .
we present our results in graphical form .
psfig # 1#2 |
one of the major pending cosmological problem for the formation of galaxies and clusters , their distribution and evolution , is the relation of the latter with the initial conditions prevailing in the early universe . to this sake ,
most of the observational constraints are relative to the luminous part of the universe . on the other hand , theoretical model of primordial fluctuations
are for the matter distribution . whether light traces matter , so that the former can be used to constrain the latter , has been a long debate over the past decade .
it was clear that from the constraint on the small - scale peculiar velocity dispersion of galaxies ( davis & peebles 1983 ) , matter can not trace light ( kaiser 1984 , davis et al .
1985 , schaeffer & silk 1985 ) if the density of the universe is to be equal to the critical density . that galaxies might be biased with respect to the matter field
has therefore been quite popular in the past years especially in the frame of the cold dark matter cosmological model .
standard cdm model provides indeed enough power at small scale for hierarchical clustering to occur but it does not not produce enough power at large scales , when correctly normalized at small scales , to explain the large scale galaxy counts ( saunders et al .
1991 , davis et al .
this problem has been confirmed by the measurement of the very large scale cosmological fluctuations by smoot et al .
( 1992 ) .
it is therefore now customary to work within the assumption of a low - density universe ( with a possible cosmological constant ) , that preserves the hierarchical clustering scenario , and with the idea that light might trace matter .
it is then crucial to understand the possible existence of biases at smaller scales and their possible variation with scale to establish reliable constraints on the shape of the power initial spectrum .
bardeen et al .
( 1986 ) proposed a mechanism for the galaxies to be more correlated that the matter .
it relies of the correlation properties of the peaks in the initial gaussian density field .
this approach was further extended by mo , jing & white ( 1997 ) .
the idea is that galaxies form at the location of the density peaks and thus are biased from the beginning compared to the whole matter field .
however such a description is far from being complete since the subsequent complex dynamical evolution of the density field is ignored . at scales
up to 8 @xmath5 the density field is completely nonlinear so that the statistical properties of the peaks appearing in the initial gaussian fluctuation may have been completely transformed .
we present below arguments that take into account the nonlinear evolution of the matter field to show that the halos in such an evolved density field indeed are distributed differently than the matter .
but _ definitely not _ in the way predicted by gaussian fluctuations .
we present the complete correlation properties that are to be expected and the expression of the biases that appear at various levels in the nonlinear regime .
the small - scale power law behavior of the galaxy correlation function is specific to hierarchical clustering scenarios and arises from the nonlinear instabilities in an expanding universe .
the value of the small - scale power law index is likely to come from the shape of the primordial power spectrum ( davis & peebles 1977 ) .
some authors , however , ( saslaw & hamilton 1984 ) propose an explanation based on thermodynamical arguments to explain the emergence of such a power law behavior . in any case
it is thought to be due of a relaxation process in the nonlinear regime . generalizing a relation of peebles ( 1980 ) , hamilton et al .
( 1991 ) propose ( see also valageas & schaeffer 1997 -hereafter vs97- ) a universal transformation to get the evolved non - linear two - point correlation function from the initial spectrum , based on empirical observations in numerical simulations .
the strength of the matter two - body correlation function is obviously an important information to derive the initial matter power spectrum , but the differences between linear regime and nonlinear regime is unlikely to be reducible solely to a transformation of the fluctuation spectrum .
indeed the observed three - point correlation function of the galaxies , for instance , also takes a specific form , as product of two two - body correlation functions ( groth & peebles 1977 ) and can provide alternative checks for the scenario of structure formation .
these features can not be predicted by analytical calculations using the simple linear approximation with initial gaussian fluctuations .
this failure is not due to inapproriate initial conditions but to the fact that the linear approximation is inadequate .
perturbative calculations introducing higher order of the overdensity field have demonstrated that the gravity can induce a full hierarchy of correlations starting with gaussian initial conditions ( fry 1984b , bernardeau 1992 , 1994 ) .
the scaling due to hierarchical clustering can be expressed through the behavior of the mean @xmath6-body connected correlation functions of the matter field within a volume @xmath7 , @xmath8 , as a function of the two - body one ( see balian & schaeffer 1989 -hereafter bas89- for a description of this statistical tools ) .
this relation can be written , @xmath9^{p-1},\eqno(1.1)\ ] ] where the coefficient @xmath10 are independent of the scale .
when the fluctuations are small ( @xmath11 ) , one can derive the full series of the coefficients @xmath10 .
unfortunately such results in the quasi - gaussian regime are irrelevant for the fully nonlinear regime where numerous shell crossings and relaxation processes have to be taken into account .
explanations for the observed structures describing the dynamics of pressure - less particles in gravitational interaction that do not assume the existence of coherent velocity flows are to be invoked .
this hierarchy of equations ( the born , bogolubov , green , kirkwood , yvon -bbgky- hierarchy ) concerns the @xmath6-body density functions ( in the full phase space ) and has been established by peebles ( 1980 ) for matter in an expanding universe .
it can not be solved in general , although there exist some interesting attempts ( hamilton 1988b , balian & schaeffer 1988 , hamilton et al .
1991 ) but admit the so - called self - similar solutions ( davis & peebles 1977 ) .
the latter solutions contain a precious indication since it can be shown that , in the limit of large fluctuations , a similar relationship as in ( 1.1 ) is expected to occur between the mean correlation functions .
models of matter fluctuations in which such a behavior is seen are called hierarchical models .
that such solutions are relevant for the evolved density field was also not obvious and had been the subject of many papers . after the success of early models ( schaeffer 1984 , 1985 ) ,
a complete description of the properties that are to be expected for the counts in cell was given by bas89 ( see 2.2.1 of this paper for a very short review ) .
these predictions have now been widely checked , for points in numerical simulations ( see valageas , lacey & schaeffer 1999 ) and in catalogues ( see benoist et al . 1998 ) .
we take now for granted that the hierarchy ( 1.1 ) holds for the analysis of galaxy catalogues , however , it is more or less assumed that the galaxy field traces the underlying density field , or at least that the properties of the matter should be preserved in the galaxy field .
the question we then have to address is to confidently relate properties concerning matter distributions to observational statistical quantities such as the two and three - point galaxy or cluster correlation functions , their cross - correlations or the galaxy and cluster void probability functions ... in the gaussian approximation , the knowledge of the @xmath6-body density functions reduces to the behavior of the two - body function . the bias ,
measuring the ratio of the strength of the two - point correlation function of the galaxies to the one of the matter is then the only relevant parameter , as stressed previously , to quantify the departure between mass fluctuations and would - be light fluctuations .
it appears to be a crucial parameter to constrain the cosmological models .
however , the actual departure between galaxy distribution and matter distribution can not be reduced to this sole parameter .
more generally we have to search how the @xmath6-body correlation functions are modified for the observed galaxy distribution . previous work ( bernardeau & schaeffer 1991 -hereafter bes91- , vs97 , valageas & schaeffer 1998 ) suggests a natural way to identify the visible objects among the matter fluctuations .
it is based on the assumption that dissipative processes are epiphenomena that do not modify the distribution of matter , but just mark , with luminous stars , the dense matter concentrations . at scales smaller than the correlation length it appears
( bas89 ) that most of the volume of the universe is ( nearly ) empty whereas a small fraction of its volume contains almost all the matter .
the latter fraction of the universe is identified with the observed luminous objects .
it happens that there is a unique parameter that governs the mass distribution function of the objects , @xmath12 where @xmath1 is the mass of the object , @xmath13 the mean mass density of the universe , @xmath7 the volume of the objects and @xmath14 the mean value of the matter two - point correlation function within a cell of volume @xmath7 .
as @xmath14 is seen to behave as @xmath15 where @xmath2 is the size of the object and @xmath3 , the power - law index of the two - body correlation function is close to 1.8 , @xmath16 scales as @xmath17 and is then nearly proportional to the internal velocity dispersion of these objects , ( which by definition are supposed to be dense enough to be virialized ) .
the main purpose of this paper is then to show how the galaxy and cluster correlations are related to the fluctuations of the matter density in the deeply nonlinear regime . in practice
, we divide the universe into cells and calculate the correlation function of those cells containing a given amount of matter ( above a rather small minimum ) . in the following
we may refer to these dense cells as `` halos '' .
all the calculations have been done for hierarchical models .
more specifically we introduced the tree - hierarchical models for which a complete description of the biases can be derived .
these models form a large part of the hierarchical models where the many - body correlation functions behave as a power of scale , and can be parameterized as products of the two - body correlation functions .
they are described in sect .
[ nonlinbe ] .
a calculation of the bias parameter was performed ( schaeffer 1985 , 1987 ) using specific tree - hierarchical based models . in a previous paper bernardeau & schaeffer ( 1992 , hereafter bes92 ) performed the calculation of the two- and three - point correlation functions of the dense spots in the matter field using a much more general form of tree - hierarchical models .
these calculations were recently extended by munshi et al .
( 1998b ) up to the 6-point functions exploiting the same techniques and models .
once again the scaling parameter @xmath16 was seen to play a central role .
the bias , the parameter @xmath18 measuring the strength of the three - point correlation function for halos , were both shown to depend only on their internal scaling parameter @xmath16 . in sect .
[ genfunc ] of this paper we present the generalization of these results to the correlations at any order .
the complete @xmath6-body correlations of condensed objects are thus derived from the matter @xmath6-body correlation functions .
[ models ] is devoted to the discussion of the various possible models , still within the tree form for the matter correlations , that can be realistically ( or less realistically ) used to describe the matter field .
the range of the prediction for the halo correlations depending on which model is adopted is discussed in great detail . in sect .
[ obscons ] we discuss various possible applications of these results , that may be checked either in galaxy catalogues or in numerical simulations .
we consider the first few moments of the counts of objects , the void probability functions .
special care is devoted to the determination of the mass multiplicity function of halos in over - dense areas ( rich clusters ) . in the last section we strike the balance of what have been obtained and present some possible developments .
the mathematical aspects of the derivations are given in the appendices .
preliminary account of this work , that contains all the major results presented here , has been given by bernardeau ( phd thesis , paris 1992 ) .
after near final completion of this paper , we learned about the overlapping work by munshi et al .
( 1998b ) .
we assume that the correlation functions in the matter field follow a hierarchical pattern as in eq .
more precisely we assume that there are solutions of the equations of motion for the dynamical evolution of the matter distribution , for which all many - body correlation functions exhibit a scaling law , @xmath19^{p-1}}\nonumber\\ & & \hmove \ \
\disp{\times\xi_p(\vr_1,\dots,\vr_p , t ) } , \eqno(2.1)\end{aligned}\ ] ] where @xmath20 is the connected @xmath6-point correlation function as a function of the @xmath6 comoving coordinates @xmath21 and time @xmath22 ( davis & peebles 1977 ) .
the scaling laws shown in this equation are valid for an einstein - de sitter universe and power - law initial conditions . for the latter , in case of @xmath23 , an other similar scaling is reached in which the coefficient @xmath24 appearing in equation ( 2.1 ) has to be changed in 1 , @xmath25^{p-1}}\nonumber\\ & & \hmove \ \
\times \disp{\xi_p(\vr_1,\dots,\vr_p , t ) } , \eqno(2.2)\end{aligned}\ ] ] which only changes the time dependence of these functions ( balian & schaeffer 1988 ) . the latter have argued that , provided either @xmath26 or @xmath27 so that the above solutions exist , they may be reached for more general initial conditions .
numerical simulations show ( davis et al .
1985 , bouchet et al . 1991 ,
colombi et al .
1994 , 1995 , 1996 , munshi et al . 1998a , valageas et al . 1999 ) that ( 2.1 ) is indeed relevant at late times , at least for @xmath26 and @xmath28 .
the matter correlation functions can not be obtained from observation , but the galaxy correlations for @xmath28 also are seen to obey eq .
( [ 2.1 ] ) or eq .
( [ 2.2 ] ) ( for the scale dependence ) .
it is thus fair to assume that either eq .
( [ 2.1 ] ) or eq . ( [ 2.2 ] ) holds also for the higher @xmath6-body correlation functions , and can be used as a model to describe the matter fluctuations in the nonlinear regime .
once the two - point correlation function is assumed to obey ( 2.1 ) or ( 2.2 ) , a further simplification can be introduced by assuming that @xmath20 can be written as a product of @xmath29 two - body correlation functions , @xmath30 where @xmath31 is a particular tree topology ( fig .
1 ) connecting the @xmath6 points without making any loop , @xmath32 is a particular labeling of the @xmath6 points by the coordinates @xmath33 , the sum corresponding to all different possible labelings for the given topology @xmath31 and the last product is made over the @xmath29 links between the @xmath6 points with two - body correlation functions .
we note for later use that , for each topology @xmath31 , we can list the number of vertices having a given number , @xmath34 , of outgoing lines , @xmath35 .
the weight @xmath36 is independent of time and is a homogeneous function of the positions , @xmath37 associated with the order of the correlation and the topology involved .
several attempts have been made using the form ( 2.3 ) .
the functions @xmath36 are generally assumed to be independent of the coordinates @xmath21 , and sometimes also , for simplicity , to be independent of @xmath38 .
for instance schaeffer ( 1984 , 1985 ) uses @xmath39 , constant or moderately increasing with @xmath6 , justified on observational grounds , so as to produce enough power towards the small masses ( power - law increase ) and a slow enough fall - off of the induced multiplicity function , which must not be faster than exponential .
this implies a very strong growth , as @xmath40 , of the @xmath6-body correlation function amplitude .
fry ( 1984a ) proposed @xmath41 , decreasing roughly like @xmath42 as a solution of the bbgky hierarchy .
this implies correlation functions that grow at most as a power of @xmath6 .
the associated mass function , for counts far from the mean , then differs by several orders of magnitude from the previous one .
this result , however , was obtained by means of approximations that were later seen ( hamilton 1988b ) to be unjustified , the accumulated error on @xmath43 being precisely of the order of @xmath40 .
the latter found an approximate solution where @xmath36 vanishes for all tree topologies , except for the particular line - topology , for which @xmath44 , indeed of moderate variation as a function of @xmath6 since @xmath45 , in this case , is of order unity .
there are , however , strong observational constraints on these coefficients due to the fact that they determine the shape of the mass function of objects ( schaeffer 1984 , 1985 , bas89 , bes91 , vs97 ) . for instance
the solution proposed by hamilton ( 1988b ) is seen to produce still too sharp a cut - off at high masses ( although exponential , as desired ) as well as at low masses , where the power - law behavior is lacking and is not realistic enough .
for this reason , bas89 made a study valid for all models that satisfy a well - established set of known constraints arising either from mathematical consistency ( such as the positivity of probabilities ) or from the requirement that the mass function should look like the schechter ( 1976 ) function .
the model proposed by schaeffer as well as the form proposed by hamilton are both a particular case of a more general form , @xmath46 where @xmath47 is a constant weight associated to a vertex of the tree topology with @xmath34 outgoing lines .
the value of @xmath36 is then obtained by the product on all the weights associated with the topology @xmath31 . the form ( 2.5 ) , assumed to hold for the matter correlations , has been used ( bes92 ) to calculate the two- and three - body correlation functions for galaxies or clusters .
we call it here the _ minimal tree - hierarchical model_. this form is supported by numerical evidences for at least the bi - spectrum which is found to be very weakly dependent on the geometry in the nonlinear regime ( see scoccimarro et al
. 1998 and references therein ) .
this is also assumed for the construction of the so - called hyperextended perturbation theory model ( scoccimarro & frieman 1999 ) .
but the aim of this article is to find the relation between the distribution of the matter and the luminous ( dense ) objects , using the most general possible form for the many - body correlation functions compatible with ( 2.3 ) .
the most general form that is relevant to the kind of calculations we develop in this paper is of the form @xmath48 the product being over all vertices associated to the tree - topology @xmath31 ; the position of the @xmath49 vertex is at @xmath50 and @xmath51 are the coordinates of the end - points of its @xmath34 outgoing lines .
the function @xmath52 is supposed to be an homogeneous function of the positions . in general
it is then supposed that the weight associated to the vertices may depend not only on the number of outgoing lines as in the minimal models , but also on the geometry of these lines such as their relative angles , or the ratios of their lengths .
this very general form of the correlations will be referred to as the _ tree hierarchical model_. if there is no definitive proof that the many - body correlation functions take the form ( 2.6 ) in the strongly nonlinear regime , this kind of structure is however obtained in the quasi - linear regime ( fry 1984b , bernardeau 1992 ) where the correlation functions follow a tree like pattern . in the following
we will mention the properties that are specifically due to the hypothesis ( 2.6 ) once the form ( 2.1 ) has been assumed . to calculate the number density of the dense and under - dense cells as a function of the mass ( , the number of matter particles ) it has been shown ( bas89 ) that it is sufficient to know the average of the many - body correlations within a cell , as is exemplified by ( 1.1 ) .
we recall here these findings , in a language that is somewhat different from the one used in these earlier publications , but that will be similar to the one relevant to the purpose of this paper .
let us divide the universe in cells of given scale , or volume @xmath7 .
the counts in cells , @xmath53 , that is the probability of finding a given amount of matter in a randomly chosen cell , can be obtained ( white 1979 , bas89 ) from a generating function that depends only on the averages , @xmath54 this implies that the counts are defined by the knowledge of a set of numbers @xmath10 , @xmath55 ( to which we add @xmath56 for convenience ) , @xmath57 which are independent of cell size and of time , provided ( 2.1 ) holds .
the counts in cells are then given by ( bas89 ) , @xmath58 where @xmath59 reads , @xmath60 we assume in this paper the coefficients @xmath10 are given .
within the minimal tree - hierarchical model ( 2.5 ) , each of the coefficients @xmath10 is related to the @xmath29 first coefficients @xmath47 ( appendix [ aappend ] ) . as a result @xmath61
can then be calculated from @xmath62 , ... @xmath63 from @xmath10 . in case of the minimal tree - hierarchical model ,
the matter correlation functions are thus entirely determined by the values of the coefficient @xmath10 , and thus by the shape of the mass distribution . for the general tree hierarchical models ,
however , the count - in - cells are seen ( a.19 ) to depend on a suitably defined average over the scale - dependent vertices @xmath47 @xmath64 so , in case only the count - in - cells are considered , the tree hierarchical model and the minimal tree hierarchical model lead to similar , and in practice , indistinguishable results .
however , a priori , the underlying correlations of the general tree hierarchical models might be quite different and more information than the count - in - cells is needed to determine all other probabilities . nevertheless , for the count - in - cells
, both models can be described in terms of constant coefficients @xmath47 .
the relation between @xmath10 and @xmath47 can be explicited through the generating function of the vertices @xmath47 , @xmath65 from which @xmath59 is easily obtained ( bes92 ) by summing ( see , jannink & des cloiseaux 1987 ) the so - called `` tree - graph series '' : @xmath66 where @xmath67 is given as a function of @xmath68 by the equation , @xmath69 the knowledge of the function @xmath59 or @xmath70 determines the multiplicity of cells as a function of their matter content .
a useful relation , that is readily derived from ( 2.12 ) and ( 2.13 ) , is @xmath71 the general properties of the cell distribution @xmath72 , gaussian or non - gaussian regime , behavior for small or large masses in the nonlinear regime can be established , provided @xmath73 ( and hence @xmath74 ) satisfies some very general constraints ( bas89 ) .
these properties , called `` model independent '' , concern any hierarchical model ( and not only the tree models considered here ) .
first , @xmath59 should be a power - law at large @xmath68 , or @xmath67 respectively : @xmath75 this implies for @xmath70 , at large @xmath67 @xmath76 secondly , the series ( 2.10 ) must have a very small radius of convergence , equal to @xmath77 .
this may be due to a rapid divergence of @xmath70 but more likely because of a singularity brought by the implicit equation ( 2.13 ) at @xmath78 , which must thus be quite close to the origin .
this calls also for a small value of the associated value @xmath79 .
the latter situation is the one encountered in the quasi - gaussian regime ( bernardeau 1992 ) .
finally , the choice ( 2.9 ) for defining @xmath16 and @xmath80 implies , for small @xmath68 and @xmath67 , that the following expansions are valid , @xmath81 @xmath82 for the dense cells , the multiplicity function in the non - linear regime , @xmath83 , takes the scaling form ( bas89 ) , @xmath84 that can be written , @xmath85 provided @xmath86 the latter constraint ensures that the matter content of the cells is large enough .
we will refer to these cells as `` halos '' .
the function @xmath87 shows that the dependences on size , density and mass only enter through the ratio @xmath88 ( see eqs .
[ [ 1.1 ] ] and [ [ 2.9 ] ] ) .
it is related to @xmath59 by @xmath89 it is defined only for @xmath16 strictly positive and behaves as @xmath90 with @xmath91 and @xmath92 with , generically , for tree - hierarchical models @xmath93 .
the mass multiplicity function thus indeed is quite similar to the usual schechter ( 1976 ) parameterization ( exponential times power - law ) of the galaxy and cluster luminosity functions .
these predictions have been successfully checked against direct galaxy counts ( alimi et al .
1990 , maurogordato et al .
1992 , benoist et al .
numerical simulations also stongly back this description of the scaling properties induced by hierarchical clustering ( bouchet et al . 1991 , colombi et al . 1994 , 1995 ) .
some departures from the scaling law however were found by colombi et al .
1996 , whereas munshi et al .
1998a find consistency with the numerical data , see discussion of this point by valageas et al .
1999 . a quite convenient form of @xmath70 , that bears all the desired feature , asymptotic form at small @xmath16 and exponential with large @xmath94 , is ( bes92 ) @xmath95 which leads to @xmath96 the choice of @xmath97 insures a good agreement with the counts in cell statistics obtained in cdm numerical simulations ( bouchet et al .
1991b ) , whereas power - law initial conditions ( colombi et al .
1996 , munshi et al .
1998b , valageas et al .
1999 ) lead to @xmath98 , for @xmath99 and , respectively , a spectral index @xmath100 .
these forms are very close to the form proposed by bernardeau ( 1992 ) derived from the equations of the dynamics in the quasi - linear approximation .
the latter finds ( 2.20 ) with @xmath101 is relevant already in the early non - linear regime ( when smoothing effects are neglected that is for @xmath102 ) .
the normalization properties for @xmath87 are worth being recalled .
indeed we have , @xmath103 reminiscent of , @xmath104 both relations are exact sum rules , showing that whenever @xmath72 in the previous integral is replaced by its approximate form , ( 2.16 ) , which is valid for the distribution of dense cells , there is no error in computing the total mass , which is thus shown to essentially be contained in these denser cells .
the normalization property , @xmath105 which means the cells occupy all the volume on the universe , however has no simple analogue for @xmath87 .
indeed we have , @xmath106 that implies that the volume fraction occupied by the dense cells described by @xmath87 is , @xmath107 which is small as compared to unity .
this shows , indeed , that @xmath87 describes the denser cells , that contain all the matter but , occupying a negligible fraction of the volume , may be expected to be surrounded by regions with a negligible fraction of the matter .
this justifies the referrence to these cells as `` halos ''
. it will be the correlations of these dense cells that we aim to calculate in the following .
clearly , would we use the full expression ( 2.9 ) for the probability , which holds in all cases , even for the nearly empty regions with @xmath108 , and include the latter in the sum , the normalization would of course be unity , eq .
( 2.22.1 ) . for completeness , we recall also that the normalization of @xmath80 ( eq . [ 2.9 ] ) that is used to define @xmath16 has been chosen so as to yield @xmath109 and more generally we have , @xmath110 important constraints can be obtained from the sum rule ( 2.22.4.1 ) .
indeed , writing that , @xmath111 that is @xmath112 whatever @xmath113 . for the value @xmath114
that minimises the above expression , we get the constraint , @xmath115 the same constraints hold if the counts are discrete , still in the limit where @xmath14 is large .
one simply has to use the exact expression ( a.2 ) of @xmath116 for the probability and the weight @xmath117 instead of @xmath118 so as to take advantage of the properties ( bas89 ) of the factorial moments .
such constraints were first derived in a slightly different form by peebles ( 1980 ) for @xmath62 and fry ( 1986 ) for all the terms .
they imply for instance , @xmath119 and also that the many - body coefficients @xmath10 can _ never vanish _ , as is directly obvious from ( 2.22.4.1 ) since @xmath87 is positive .
there are also several mathematical forms for @xmath87 that will be useful . integrating ( 2.18 ) by parts ,
we get , @xmath120 using @xmath67 as a variable , a second integration by parts gives , @xmath121 where however the integration contour of @xmath67 has to be suitably defined according to the function @xmath122 defined by ( 2.13 ) .
we still consider the universe divided in small cells . the central problem
we solve in this paper is the calculation of the correlations , at any order , among cells that contain a specific amount of matter .
we consider a large volume @xmath123 that contains @xmath124 cells , labeled by @xmath125 = @xmath126 , of volume @xmath127 , respectively .
our aim is to derive the joint mass distribution functions @xmath128 , probability that the @xmath124 cells contain respectively the masses @xmath129 .
we will then assume that such a joint distribution function is the joint density distribution functions of the halos of the density field .
we assign a position @xmath50 to each of the cells , to which we associate respectively the typical masses @xmath130 and @xmath131 defined as previously .
we then restrict our calculation to the case @xmath132 which corresponds to the condensed objects .
the case of the joint mass distribution function for two cells has already been addressed in a previous paper ( bes92 ) .
we have @xmath133 when @xmath134 , so that the scaling ( 2.16 ) applies , the function @xmath135 is related to a new series of parameter @xmath136 , that instead of being averages over a unique cell ( 2.8 ) are now defined as mixed averages within two ( small ) cells of volume @xmath137 and @xmath138 , respectively , under the constraint that the center @xmath139 and @xmath140 of these two ( distant ) cells remain fixed . in general , for tree models , the @xmath136 parameters depend only on the average of the @xmath141-point correlation function when @xmath6 points are in one cell and the other in the second .
a similar situation is encountered in the quasi - linear regime ( bernardeau 1996 ) .
we thus have generically , @xmath142 in the case of the hierarchical models , the @xmath136 parameters are independent of both the cell volume and @xmath143 ( as soon as the finite cell size effect are neglected , , @xmath144 ( resp .
@xmath145 ) @xmath146 ) and are specific numbers related to the matter correlation functions .
the function @xmath135 is then related to @xmath136 exactly by the same way @xmath72 is related to @xmath10 .
we can as well defined the generating function of this new series @xmath147 which permits the calculation of @xmath135 , @xmath148 in case of the minimal tree - hierarchical model we have ( bes92 and appendix [ bappend ] ) , @xmath149 where @xmath74 is given by ( 2.11 ) and @xmath67 is solution of the equation ( 2.13 ) , so that we have the simple relation @xmath150 . for the general tree - hierarchical models the situation is slightly more complicated and the @xmath136 coefficients depend on vertices @xmath47 . with outgoing lines starting at @xmath151 , one ending at @xmath152 ( far away ) outside ,
whereas @xmath153 ones end within the cell centered at @xmath154 , at @xmath155 that are averaged over the cell volume , in a procedure similar to ( 2.24 ) , we get , @xmath156 we can define the generating function for these vertex weights ( @xmath157 ) , @xmath158 the function @xmath159 is then related to the function @xmath160 by the relationship , @xmath161 this equation defines @xmath159 in terms of the function @xmath122 given by ( 2.13 ) which depends only on the statistics within one cell and is thus considered here to be known .
the problem is then reduced to the construction of @xmath160 , that is to calculate the averages ( 2.27 ) . for the minimal tree models where @xmath47 is constant
, this is straightforward and the quantities @xmath162 just identify to the mere vertices @xmath47 , so we have @xmath163 .
the relation ( 2.23 ) together with the function @xmath159 gives the correlation function between two cells , @xmath164 which has the same spatial dependence than the matter with a bias factor @xmath165 given by , @xmath166 this function describes the departure between the matter correlation and the halo correlation for the two - point function .
it turns out to be a function of @xmath167 only , @xmath168 so that ( 2.23 ) reads , @xmath169 .
\eqno(2.33)\end{aligned}\ ] ] where @xmath170 ( resp .
@xmath171 ) is the internal @xmath16 parameter of cell 1 ( resp .
two _ model independent _ properties ( in the sense they do not depend on the function @xmath160 to be constructed in a specific model ) result thus from hierarchical clustering in the non - linear regime : a _ factorization property _ , the total bias being a product of two factors that refer to the internal properties of each cell , and a _ scaling property _ , these factors depending only on the internal scaling parameter @xmath16 of each cell .
the observational consequences of these results have been widely discussed in bes92 .
these predictions have been checked against observations .
benoist et al .
( 1996 ) show that the amplitude of the galaxy correlation function is definitely dependent on their luminosity ( the existence of a luminosity dependence of the galaxy correlations showing their distribution is biased with respect to the matter was by no means obvious : see hamilton ( 1988a ) and valls - gabaud et al .
( 1989 ) for an early discussion on this point ) . with the uncertainty related to the transformation of our @xmath16 parameter into luminosity
( see bes91 , vs98 ) , the observed bias follows very closely the predicted trend .
numerical simulations also ( munshi et al . 1999 ) exhibit the predicted scaling of @xmath172 with @xmath16 as well as the factorization property ( 2.33 ) .
the above considerations can be generalized to calculate the correlations of an arbitrary number of cells , again under the condition that their distance is much larger than their size .
in general , for a given number of cells the connected part of the joint mass distribution function follows the formation rules of the tree formalism ( the demonstrations are given in detail in the appendices ) .
indeed the result is , @xmath173 with the same structure as the one encountered in equation ( 2.3 ) .
the normalization parameters @xmath174 are homogeneous functions of the geometry of the positions of the cells @xmath175 , and functions of the masses @xmath176 attributed to each of the cells . for each particular tree connecting the @xmath124 cells ,
@xmath177 is obtained by a product over the set of vertices ( whose order we label by @xmath178 ) associated to the tree @xmath31 , @xmath179 this form of @xmath180 defines `` effective '' vertices @xmath181 , @xmath182 the vertex associated with the cell labeled @xmath183 depends on the mass @xmath184 and possibly on the geometry of the @xmath178 outgoing lines that appear in the tree representation of each particular term . the functions @xmath185 are related to the matter correlation function properties .
more specifically these functions are generated by an @xmath186 series , defined by , @xmath187 in case of the minimal tree model , the parameters @xmath186 are pure numbers that do not depend on the size of the cells ( @xmath127 ) , neither on the geometries of the @xmath178 outgoing lines .
for the general hierarchical models a dependence with the geometry of these lines is expected since the matter correlation function are supposed to present such a dependence .
the derivation of the value associated to the dressed vertices @xmath185 is based on the same principle as for @xmath72 and @xmath135 , @xmath188 with @xmath189 as can be seen by comparing ( 2.6 ) and ( 2.34a ) , the dressed @xmath174 can then be written as a product of vertices given by @xmath190 we have obviously @xmath191 and @xmath192.the correlation functions of the condensed objects are thus based on a tree structure similar to the one of the matter . as the distribution of the matter field is given by the `` microscopic '' vertices @xmath47 , that depend on position unless we use the minimal tree hierarchical model , the distribution of the cells is given by these new `` dressed '' vertices @xmath193 .
these vertices depend on the internal cell properties through the scaling parameter @xmath16 of the cell .
they do not depend on position because all the calculations are done in the limit where the cell size is much smaller than the distance between cells . in the following the relationship between @xmath194 and the matter correlation functions will be explicited .
we naturally expect that these results obtained for cells randomly distributed in the universe are also valid for the counts of astrophysical objects
. a precise calculation of the halo correlations would require a more complicated calculation , but the general properties we find here are expected to be valid .
some departures can not naturally be excluded although they are expected to be small since the densest spots of the present universe contrast extremely well with the matter field . before calculating the vertices @xmath194
, we can already derive a set of general properties of the dressed correlation functions , that are independent of the specific model used for the matter correlation functions .
they depend only on the assumption of an underlying tree structure of the latter .
these results are the core of what has to be retained from these calculations .
first of all in the strongly nonlinear regime the correlation functions of the halos have the same scale , and time , dependence as the matter : @xmath195^{p-1 } \xi_p^{\ob}(\vr_1,\dots,\vr_p , t ) } , \eqno(2.39a)\end{aligned}\ ] ] and so that the hierarchical properties assumed to the matter field should also be present in the galaxy field , or in the cluster field . as a result the ratios , @xmath196 are also scale independent .
the fact that indeed the observations confirm the existence of such property is an indication of the validity of the hierarchical hypothesis at the matter distribution level .
it is thus a confirmation of the validity of the hypothesis ( 2.6 ) .
the index @xmath3 measuring the slope of the two - point galaxy correlation function in the nonlinear regime has also to remain identical to the one of the matter field , although the normalizations are not the same ( see bes92 for a complete discussion ) . as a result the pair velocity correlation between galaxies is expected to have the distance dependence expected in the hypothesis of no - bias .
this is what is indeed observed as stressed by peebles ( 1987 ) , and is by no means in contradiction the existence of a non - trivial natural bias .
however it does not mean that the large scale ( quasilinear , see bernardeau 1996 ) and small scale ( nonlinear , considered here ) biases should be the same .
a short examination of the equation ( 2.27 ) proves that the parameters involved for the values of @xmath172 are completely determined by the nonlinear regime in case of the small - scale bias , whereas they are given by a mixture of the small scale and large scale correlation function behaviors for the bias at large scale . a quantitative ,
if not qualitative , change is expected to occur when one goes from one regime to the other , which may lead to a variation of @xmath197 .
this is not in contradiction with the previous remark on the index @xmath3 since in the former case , we suppose to be only in the strongly nonlinear regime . the normalization parameters ( the @xmath198 in the equation [ [ 2.39b ] ] ) that are measurable for the galaxy or cluster fields have no reason to be identical to the ones of the matter field .
as a result it would be a complete nonsense to try to discriminate theories with the use , for instance , of the skewness of the matter field obtained in numerical simulation with the one of the galaxy field .
the strength of the correlation functions , so the biases at any level , depend only on the internal parameter , @xmath16 , of the objects .
this is a property of great interest for observational checks
. indeed objects of different natures can have the same @xmath16 parameter .
for instance the brightest galaxies and the rich clusters should have the same biases , the `` common '' galaxies should have biases comparable to ones of the groups ( bes92 ) .
such features can easily be checked in the present or coming catalogues .
the last property concerns the @xmath194 parameters that govern the correlation strength .
it has just been stressed that they are a function of @xmath16 only .
we also point out that they describe as well the auto - correlation functions as the cross - correlation functions between objects of different kind or between objects and the matter field .
what does it mean in practice ?
consider for instance three fields of any kind , galaxy , cluster or matter
. then one may want to know what is the three - point density function of these fields , , what is the joint expected densities of these three fields .
the three - point density function @xmath199 can be written @xmath200 where @xmath201 is the density of the objects of kind 1 , @xmath202 of kind 2 and @xmath203 of kind 3 , @xmath204 , @xmath205 and @xmath206 are their respective bias parameters .
the coefficients @xmath207 are the parameters describing the three - point correlation function of the objects of kind @xmath125 . the possible geometric dependence in fig .
( 2 ) we give a diagrammatic representation of this relation .
these kinds of relations based on tree constructions can be done at any order where the correlation properties of any kind of object is fully determined by the series of the vertices @xmath193 .
the knowledge of the vertices @xmath193 then give very diverse results concerning the auto - correlation properties as well as the cross - correlation properties of any kind .
this may be of great interest since one can imagine observational tests to check such a property .
the simplest is to check that the auto - correlation function of galaxies , @xmath208 , the cross - correlation between galaxies and clusters , @xmath209 , and the auto - correlation function of clusters , @xmath210 , verify the relation , @xmath211 one can also construct some relations involving the three - point correlation functions between galaxies and clusters that should be verified in catalogues , @xmath212 the verification of these relations is a check of the hypothesis ( 2.1 ) as well as of the hypothesis ( 2.6 ) .
it would also be the indication that the clustering properties of the astrophysical objects have a pure gravitational origin .
the relation ( 2.40 ) , in case the objects are of the same kind , reduces to , @xmath213 , \eqno(2.42)\end{aligned}\ ] ] where @xmath214 is the measurable two - point correlation function and @xmath215 .
the latter is the parameter describing the three - point correlation function and can be measured directly .
such a remark can be made at any order so that we are led to define effective vertex weights , @xmath216 ( note that @xmath217 ) . these parameters are the only ones that can be measured directly in catalogues and the only quantity that can not be measured directly is @xmath197 . as it has been discussed previously the correlation functions of the astrophysical objects are completely determined by the parameters @xmath194 ( eq . [ [ 2.38 ] ] ) .
the latter are determined by the parameters @xmath186 ( eq . [ [ 2.35 ] ] ) that depend on the vertices of the matter correlation functions .
they depend more precisely on averages over the elementary vertices @xmath47 , located at @xmath218 in @xmath123 , with @xmath34 outgoing lines , @xmath178 of which end up in @xmath123 far from @xmath218 and @xmath219 of which end within a volume @xmath7 centered at @xmath218 @xmath220 where @xmath34 represents the total number of lines connected to the vertex and @xmath178 is the number of long lines ( corresponding to the volume @xmath123 ) . in the tree models
@xmath221 is a parameter independent of the size of the volumes @xmath7 and @xmath123 . in case of the minimal tree model ,
one obtains @xmath222 but we prefer to work without such a restrictive hypothesis .
these parameters define a function of two variables , @xmath223 in case of the minimal tree model the function @xmath224 is simply a function of @xmath225 and can be directly deduced from the knowledge of the count - in - cells statistics , that is the knowledge of the function @xmath74 , @xmath226 in general , all the informations that are required for the calculations of the many - body correlation function bias we want to do are contained in the function @xmath224 . on the other hand
the halo correlations are entirely driven by the parameters @xmath194 that , as for the matter field , define a generating function @xmath227 as a result , the halos of the density field of scaling parameter @xmath16 have correlations given by the function @xmath228 with the meaning of the equations ( [ 2.3]-[2.6 ] ) . in the appendices ,
we derive the expression of @xmath228 from the expression of @xmath224 .
the result can be expressed in terms of an auxiliary function , @xmath229 , through , @xmath230 the function @xmath229 generalizes the function @xmath87 with an extra dependence with the parameter @xmath231 that describes the environment .
we have , @xmath232 we note here that , since @xmath233 , we have @xmath234 and @xmath235 ( this justifies the notation we adopted ) .
this is the central result of this paper .
it is established in detail in the appendix [ bappend ] .
it is valid only provided ( [ 2.17 ] ) is satisfied so as the scaling in @xmath16 holds , as we generically assume here for the objects we call halos .
the expression for the vertex generating function in the general case , when the lower masses are included , and even in case the density field is represented by discrete points , is given in app .
[ bappend ] .
it amounts in general to use @xmath236 instead of @xmath237 in eq .
( [ 2.47b ] ) , the definition of @xmath74 , eq . ( [ 2.47a ] ) , being changed accordingly . in this case
there is no scaling in @xmath16 for the underdense cells , and the @xmath16-dependence is to be replaced by the dependence on mass @xmath1 .
the equations ( [ 2.47a]-[2.47d ] ) explicit the relationship , @xmath238 between @xmath239 and the vertices @xmath221 that are directly derived from the matter correlations as a relation between @xmath239 and @xmath224 . from @xmath240 , it is then possible to deduce @xmath228 .
the latter may be directly related to @xmath241 by means of eqs .
( 2.47.2 ) or ( 2.47.3 ) .
the vertex generating function @xmath228 plays the same role for describing the statistics of the cells labeled by their parameter @xmath16 as the generating function @xmath70 which gives the statistics of the matter field ( [ 2.9 ] , [ 2.12 ] ) .
the expansion of @xmath240 in powers of @xmath231 defines the dressed @xmath242 coefficients , whose equations can also be obtained by directly expanding ( [ 2.47a]-[2.47d ] ) .
all properties of the halo correlations can be derived from this equation . in particular
it is easy to see that , @xmath243 which is the generating function of the trees with one _ long _ external line .
these relations can be written in several equivalent forms . from , @xmath244 which is analogous to ( 2.13.1 ) , eq .
( 2.47b ) can also be written @xmath245 as well as , after a second integration by parts , @xmath246 as in ( 2.22.6 ) , the integration is to be made over a contour that is defined by the relation ( 2.47d ) which determines the function @xmath247 . considering the function @xmath228 defined in equation ( 2.46 ) and given in ( [ 2.47a]-[2.47d ] ) it is straightforward to demonstrate that
, @xmath248 which means that whatever @xmath178 we have , @xmath249 these normalization properties ( eqs . [ [ 2.48 ] , [ 2.49 ] ] ) are quite important and deserve attention .
it means that at a given scale _ the mean correlations of the dense cells are the ones of the matter field _
, once the mean is calculated in a proper way : each object has to be weighted be its own internal scaling parameter @xmath16 which is equivalent to have the probability weighted by the mass .
it is thus clearly seen that the dense cells represent the matter content of the universe : when properly averaged , the dense cell correlations are identical to the correlations of the matter points , at all orders .
it implies that the halos are not biased as a whole compared to the matter field .
nevertheless , a population of particular object is always biased because there is a segregation in the correlation properties of the halos , @xmath194 varies with @xmath16 .
it is clear from ( 2.49 ) that an @xmath16-independent vertex @xmath250 implies @xmath251 .
the fact that the clusters are more correlated that the galaxies ( bahcall 1979 ) or a luminosity - dependence of the galaxy correlation function ( benoist et al .
1998 ) is an indication that such a segregation indeed exists in the universe and implies that galaxies are not expected to be correlated as the matter field .
another sum rule involves also the underdense cells ( app . [ bappend ] ) , and in the continuous limit ( where the @xmath16-dependence of the vertices is to be replaced by their @xmath1-dependence ) reads @xmath252 which means that whatever @xmath178 we have , @xmath253 so , _ the underdense cells must be necessarily antibiased _ ( @xmath254 ) , as a generalisation of the antibiaising property found ( bes92 ) for @xmath165 ( @xmath255 here ) that was seen to change sign for @xmath256 .
similar relations can be written for a distribution of points , and are given in app .
[ bappend ] .
the calculations that are presented in the previous section are valid under three conditions : * the spatial size of the halo has to be smaller that the correlation length so that the mean value of the two - point matter correlation function in the volume occupied by the object is larger than unity . *
the space dependence of the correlation functions is valid when the size of the halos is smaller than their ( mean ) relative distances .
more precisely the calculations are made to leading order in the limit of vanishing values of @xmath257 where @xmath258 is the mean value of the matter correlation function within the object @xmath125 and @xmath259 is its mean value at a scale at which we want to derive the halo correlations .
for instance if galaxies are assumed to have a size of @xmath260 including their dark halo , at the @xmath261 scale the corrective terms corresponding to finite cell size effects for their correlation functions are expected to be of the order of 1% . *
the scale dependence in @xmath16 is valid when the internal mass exceeds a mass threshold @xmath262 depending on the size of the object . at the @xmath260 scale
we have @xmath263 so that galaxies are far above that limit .
the only parameters that can be directly measured in the catalogues are the parameters @xmath264 as defined in equation ( 2.43 ) .
their generating function , @xmath265 is related to @xmath228 by @xmath266 the main analytical result that can be derived from the equation ( [ 2.47a]-[2.47d ] ) concerns the large @xmath16 behavior . with the sole assumption that @xmath224 is a regular function that does not contain any spurious singularity for the useful corresponds to a singularity of implicit equation ( 2.47d ) for @xmath67 rather than to a singularity of @xmath74 and hence to a finite value of @xmath79 around which @xmath74 can be expanded , as is required by the model - independent condition , sect .
[ modindprop ] . the required condition
are not fulfilled , for instance , by the model of hamilton where @xmath79 is infinite . ]
values of @xmath67 , we obtain a general result that is valid for any tree - hierarchical model ( appendix [ dappend ] ) , @xmath267 when @xmath268 . note that this limit concerns fully nonlinear objects but is valid for any correlation scale ( linear or nonlinear ) .
this result has been presented in an early work ( bernardeau , phd thesis , paris 1992 ) , and is suggested by the results obtained by munshi et al .
( 1998b ) .
it is relevant a priori for describing the correlation of very bright galaxies .
indeed , the observations called immediately at the time these concepts were discovered for such a modeling of the distribution of bright galaxies : the @xmath269 model of schaeffer ( 1984 , 1985 ) for the galaxy many - body correlations is equivalent to @xmath270 in the language of the present paper .
we are now all set to give , as one of the possible applications of the formulae established previously , the statistics of the number of cells of size @xmath7 within volume @xmath123 that contain a mass larger than @xmath1 , ( which is labeled by a scaling parameter larger than @xmath16 ) .
for simplicity we call these cells `` full '' cells .
the probability @xmath271 to have @xmath272 of such full cells can be written @xmath273 with ( app . [ bappend ] ) , @xmath274 where @xmath275 is the average number density of full cells .
this is the analogue of ( 2.9 ) for discrete counts .
for @xmath272 large enough we have @xmath276 with @xmath277 corresponding exactly to ( 2.9 ) .
the moments of @xmath278 then are by construction @xmath279 the positivity of @xmath280 then implies the positivity of the @xmath198 coefficients as well as the constraints ( 2.22.4.4 ) for the latter .
identical formulae can be written for halos containing a mass between @xmath1 and @xmath281 , using @xmath282 in place of @xmath283 and @xmath284 in place of @xmath285 , with the same constraints ( 2.22.4.4 ) on @xmath286 .
as can be seen in the equation ( 2.43 ) the form of the halo correlation functions is entirely determined by @xmath224 .
this function is expected to obey to various rules . for the general case ,
two scales are involved .
when both are in the nonlinear regime , one expects that , @xmath287 the numerical results presented in the following part have then been partly obtained with , @xmath288^{-\ka } \ \ \ { \rm with}\ \ \
\ka = 1.3 , \eqno(2.55)\ ] ] relevant ( bes92 ) to the non - linear regime with cdm initial conditions .
however , we do not exclude that the larger scale may reach the linear regime .
in such a case the relation ( 2.54 ) is not expected to be valid , but the parameters @xmath289 are expected to take the values found in the quasi - gaussian regime from perturbative calculations ( bernardeau 1992 , 1994 ) that is for instance @xmath290^{-3/2}$ ] independently of @xmath291 for @xmath102 .
we also introduce an other explicit model that does not lead to mathematical inconsistencies .
this one is based on a property of `` scale factorizability '' which means that we assume that @xmath292 .
this can be understood qualitatively since we have two scales @xmath7 and @xmath123 different enough , so that the averages are in some sense independent . in such a case
the function @xmath224 is factorized and @xmath293 we can then adopt the form , @xmath294^{-\ka } \left[1+{\theta\over \ka}\right]^{-\ka}. \eqno(2.56)\ ] ] in both cases the function @xmath295 is known to agree with all the constraints on @xmath87 ( bes92 ) .
the form ( 2.56 ) is in agreement with equation ( 2.54 ) .
the numerical results presented in the following have been obtained with these two forms .
we will also present `` model - independent '' predictions , obtained by the mere assumption , @xmath296 for large @xmath67 , with , @xmath297 this leaves out the case where @xmath70 is a polynomial in @xmath67 , that will be discussed below as a generalization of the hamilton ( 1988b ) model .
the minimal tree model is a priori an attractive model , for which the geometrical dependence of the matter correlation functions are well defined and entirely determined by the mass function .
it is thus natural to try to calculate the cell correlations in such a model .
the function @xmath59 for the dressed vertices in the minimal tree - hierarchical model , given by using , @xmath298 as a new function depending on @xmath68 and @xmath231 through , @xmath299 can be written , @xmath300 successive derivatives of @xmath240 with respect to @xmath231 will give @xmath301 as an application of these results we have , @xmath302 where the derivatives are with respect to @xmath22 , now obviously related to @xmath68 by ( [ tdey ] ) with @xmath303 , @xmath304 the results for @xmath305 and @xmath306 have been given in bes92 and the three others by munshi et al .
( 1998b ) from an order by order expansion of the tree summation . the general expression ( [ phiqmin ] ) obtained here allow a simple and direct calculation of these functions to an arbitrary order-[2.47d ] ) , these expressions are valid only provided ( [ 2.17 ] ) holds , as we assume throughout for what we call halos . in general ( see app .
[ bappend ] ) , @xmath307 in eqs .
( [ phi1min]-[phiqmin ] ) is to be replaced by @xmath308 $ ] . ] .
it is also possible to express @xmath229 directly as a function of @xmath309 @xmath310 this equation can be written , @xmath311 with , @xmath312 this form is well suited for an expansion in powers of @xmath231 and allows one to get directly an expression for the vertices @xmath194 . with the definition @xmath313^{1-\q } \ , \eqno(e.6)\ ] ] we get @xmath193 in terms of @xmath74 , @xmath314 from the latter expression , it is readily seen that the functions @xmath315 can be formally expressed as @xmath316 in the limit @xmath317 , @xmath68 gets of order @xmath318 and @xmath22 as well gets very large .
this immediately shows that in the latter limit , only the _ model - independent _ asymptotic form ( 2.15 ) of , @xmath319 will contribute .
our calculation , thus , will be valid for any minimal tree hierarchical model satisfying the `` model - independent '' conditions . the solution of ( e.4 ) then is @xmath320 with @xmath321 this leads to the expression of @xmath322 and we get , @xmath323 \over \gamma \left[-(1-\alpha)(\q-2)\right ] } ( \ka c y)^{-\alpha(\q-2)}\ . \eqno(e.12)\ ] ] with the help of the relation @xmath324 the dressed vertices in the small @xmath16 limit can then be expressed as @xmath325 \over \gamma \left[-(1-\alpha)(\q-2)\right ] } x^{\alpha \q}\ .
\eqno(e.14)\ ] ] we see that @xmath326 \over \gamma \left[1-\alpha\right ] } x^{\alpha}\ , \eqno(e.15)\ ] ] which is the bias calculated in bes92 .
also @xmath327 that vanishes to leading order in @xmath16 and is given by higher terms than the one arising from the model - independent form ( e.9 ) , as discussed by the latter authors .
this violates the lower bound ( 2.22.4.5 ) on @xmath18 ( peebles 1980 ) , now applied to the moments of ( 2.53.9 ) that are directly related to @xmath194 , and implies that some of the values of the associated probabilities are negative in the small @xmath16 limit .
this problem can not be cured by including the terms , eq .
( [ nu2compl ] ) , that arise for @xmath328 since it is always possible to choose the volume @xmath7 small enough so as to have @xmath329 arbitrarily small .
note however that the generating function of , @xmath330 is @xmath331 , so the corresponding vertices are @xmath332 \over ( 2-\q)\gamma \left[(1-\alpha)(2-\q)\right ] } x^{\alpha \q}\ .
\eqno(e.19)\ ] ] the corresponding @xmath333 parameters , the first few are plotted on figs [ s3]-[s6 ] , are all finite at small @xmath16 and do not violate the positivity constraints ( 2.22.4.4 ) . in this case
it is actually possible to get the generating function for the vertices , @xmath334 with @xmath335\over \gamma\left[\kappa/2+\kappa\right]}\ ] ] so that @xmath336 .
when @xmath337 it has an asymptotic behavior given by , @xmath338 \over\gamma\left[(\kappa-1)/(\kappa+1)\right]}\,\beta^{-2/(\kappa+1 ) } \times\nonumber\\ & & \hmove\ \
\times\theta^{-2/(\kappa+1)}.\end{aligned}\ ] ] it also corresponds to moments @xmath286 that satisfy the constraints ( 2.22.4.4 ) .
we show in appendix that for all `` model - independent '' formulations ( that is all models satisfying a series of needed constraints related to the count - in - cells , see bas89 and discussion in sect .
[ modindprop ] ) , some of the higher moments of the cell correlations become unduly negative at small @xmath16 ( see figs .
[ s4]-[s6 ] ) .
this indicates clearly that such a model for the matter distribution is mathematically unsafe .
since there is no problem for the larger values of @xmath16 , we prefer to say that the extension of the minimal model to calculate the cell correlations is insufficient . but the calculation of the higher orders of perturbation theory may actually not be necessary : small deviations from the minimal model also may be sufficient to cure the problem .
such studies , however , are left for future work . ] to describe the distribution of the less dense among the over - dense cells .
the form found for @xmath172 and @xmath339 for @xmath340 are attractive enough to explore more completely the consequences of such a model on the whole galaxy clustering properties .
hamilton ( 1988b ) proposed another form for the vertex generating function @xmath74 of the minimal tree - hierarchical model , that does not enter into the class that we have called `` model - independent '' .
the latter obey conditions to be necessarily fulfilled so as to reproduce the known properties of the galaxy and matter count - in - cell distribution .
we have however seen in the previous section that lead to difficulties for the minimal model .
the question then may be raised whether by relaxing the model - independent conditions , probability laws that are acceptable for all values of @xmath16 can be found .
the vertex generating function of hamilton ( 1988b ) , in its original form , is , @xmath341 with arbitrary @xmath342 .
the solution of the equation ( 2.13 ) relating the @xmath67 to the @xmath68 parameter is , @xmath343 and @xmath59 is , according to ( 2.12 ) , @xmath344 from the requirement for @xmath59 to be positive for all @xmath345 ( see discussion in bas89 ) , and specifically here for large @xmath68 , we get the constraint , @xmath346 the corresponding function @xmath87 ( for @xmath16 strictly positive ) is then ( eq . [ 2.18 ] )
, @xmath347 it is readily seen that , unless @xmath348 , this form does not satisfy the normalization condition ( 2.21 ) and ( 2.22.4 ) .
nevertheless , the form ( e.25 ) is perfectly suited to our discussion for @xmath349 .
simply there is only a fraction @xmath350 of the matter in the dense cells , the remainder being in the underdense cells , whereas in the generic case the latter is negligible .
the form ( e.25 ) , because of its restriction to @xmath351 amounts to write @xmath59 , @xmath352 and to note that the last term does not contribute to @xmath87 .
if , however , we want to exclude scaling function with such peculiarities , we are restricted to , @xmath353 a case that was specially discussed by hamilton since it allows one to get many - body correlations with a structure that repeats itself under the grouping of a subset of points within an infinitesimal cell . in the case ( e.26 ) , the scaling function is , @xmath354 this is an important qualitative difference . in the actual counts , as discussed in bas89 and confirmed by observations ( benoist et al .
1998 and references therein ) as well as simulations ( valageas et al . 1999 and references therein ) , @xmath355 diverges for @xmath356 , which means that the number of small objects increases without limit , so that their total number is not defined : this means that in realistic models the total number of objects depends on _ other parameters than the ones introduced for the count - in - cells_. for the models considered here on the other hand @xmath357 is finite and the total number of objects is determined by the same physics that the dense cell .
so , the models considered in this section are too different from these realistic models to be used as a modeling of the matter distribution .
the equation ( 2.47d ) with the form ( e.21 ) for @xmath74 has the solution , @xmath358 and ( 2.47c ) yields , @xmath359 the last , constant , term does not contribute to the integral ( 2.47b ) that defines @xmath228 and we have , @xmath360 so , the model ( e.21 ) corresponds to a constant bias , @xmath361 normalized to unity since @xmath362 is normalized to @xmath350 . replacing @xmath231 by @xmath363
, we get the vertex generating function for the dense cells , @xmath364 this is nothing but the @xmath365 model .
so , for an initial generating function with @xmath365 , there is _ no bias _ for the distribution of the halos .
hamilton s condition for the correlation functions to repeat themselves when changing scale is equivalent to have an unbiased distribution . for @xmath366 ,
the distribution of cells of size @xmath7 at the larger scale @xmath123 goes over to a @xmath365 distribution .
let us consider here the general case where @xmath367 has a zero for finite values of @xmath22 @xmath368 around @xmath369 , @xmath74 has the expansion @xmath370 if we transform these relations into equalities , for @xmath371 , @xmath372 , this is the hamilton model .
the equation ( 2.13 ) for @xmath59 has , in the vicinity of @xmath373 , the solution , @xmath374 so , the behavior of @xmath74 at @xmath369 is related to the behavior at large @xmath68 of @xmath73 . that , according to ( 2.12 ) , has the expansion , @xmath375 this yields the behavior of @xmath87 at small @xmath16 , @xmath376 then with the usual definition ( [ 2.47a]-[2.47d ] ) of @xmath68 for the correlated cells and the definition ( e.1 ) for @xmath22 the equation ( 2.47c ) reads , @xmath377 and has the solution , @xmath378 this yields the expansion , at large @xmath68 , @xmath379 we then get for @xmath229 , @xmath380 the vertex generating function for the cell correlations , in the limit @xmath317 is then @xmath381 its derivative has a zero of order @xmath382 at @xmath383 @xmath384 clearly none of these forms enter in the `` model - independent '' category defined in sect .
[ modindprop ] since @xmath87 has not the required behavior at the origin . in this paragraph
we explore the consequences of the hypothesis ( 2.55.1 ) on the generating function @xmath224 .
we will see that this `` scale - factoraziblity '' hypothesis will provide us with a phenomenological model that has all the desired properties . with the assumption ( 2.55.1 ) , from ( [ 2.47a]-[2.47d ] )
, it is readily seen that , @xmath385\ , \eqno(e.45)\ ] ] with , @xmath386 this implies the simple result , @xmath387 ) \ , \eqno(e.47)\ ] ] we thus get , @xmath388 \
, \eqno(e.48)\ ] ] and @xmath389\ .\ ] ] it leads to , @xmath390 \ , \\
\zetax(>x,\theta ) & = & { 1\over h(>x ) } h\left[>{x\over\zeta(\theta)}\right ] \ .
\eqno(e.49 ) \end{aligned}\ ] ] a systematic expansion in powers of @xmath231 gives @xmath391 and @xmath392 the first of these equations implies an @xmath16-dependent bias that is expressed in terms of the scaling function @xmath87 as @xmath393 note also that , @xmath394 for large @xmath16 , from ( 2.19 ) , the bias is @xmath395 with generically @xmath396 . with the form ( e.53 ) , it is readily seen from ( e.49 ) that @xmath397 \sim e^{-\theta } \ . \eqno(e.54)\ ] ] we have the same limiting form for @xmath398 . at @xmath317
, we have , @xmath399 which , with @xmath400 , gives @xmath401 the same results are obtained for @xmath402 and @xmath403 .
finally for @xmath404 we have @xmath405 with @xmath406 interestingly , for @xmath407 or for @xmath408 , @xmath409 is invariant under the transformation of the correlations of elementary points into the cell correlations . in this case
we have indeed , @xmath410 note however that this form is not preserved for any intermediate value of @xmath16 .
the results presented throughout this paper give the qualitative and quantitative behavior of the halo correlations in the evolved nonlinear density field .
they are based on assumptions on the nonlinear matter correlation functions that have been now widely checked . nevertheless in this section
we aim to discuss the various results that we obtain with regards to the hypothesis that have been made at several stages .
the shape of the nonlinear mass distribution function is derived from the assumption that the correlation functions follow a hierarchical pattern ( 2.8 ) .
this form implies that the mass distribution function takes the form ( 2.16 ) .
the joint mass distribution function for two cells is determined by the behavior of the correlation functions when the points separate into two different subsets that lie in two volumes @xmath137 and @xmath138 at the positions @xmath154 and @xmath152 . for any hierarchical clustering
we obtain @xmath411^{p_1 - 1}\ \xi_2(\vr_1,\vr_2 ) \ \left[\xib_2(v_2)\right]^{p_2 - 1 } , \eqno(2.57)\end{aligned}\ ] ] which implies that @xmath412 . \eqno(2.58)\end{aligned}\ ] ] the resulting correlation between the halos at the position @xmath154 and @xmath152 exhibits some of the general properties previously mentioned : * the distance dependence is the same as the one of the matter correlation function ; * the bias parameter is a function of @xmath16 only .
but at this level of hypothesis there is no property of factorizability .
this latter property is obtained with the tree hypothesis which implies that @xmath413 . as a result
the bias function can be factorized since , @xmath414 which under the tree - hierarchical hypothesis reads @xmath415 such remarks can be generalized to any order of the halo correlations . as a result
we get the following properties : the hierarchical hypothesis implies that * the halos are correlated in a hierarchical way ( eq . [ [ 2.39b ] ] ) with the same scale dependence as the matter correlation functions ; * the strength of the correlation functions , whatever the order , is a function of the scaling parameter @xmath416 only . the _ tree - hierarchical _ hypothesis implies that * the halo correlations follow a tree - hierarchical pattern too ; * they take a specific form ( eq . [ [ 2.53 ] ] ) in the very massive halo limit .
note that in case of the minimal tree model , the halo correlations also follow a minimal tree - hierarchical model .
the results obtained for the two - point correlation functions of the halos have already been discussed on an observational point of view in a previous paper ( bes92 ) . in that paper
we only gave results for the minimal tree - hierarchical model .
in this paper we introduce an other specific model ( eq . [ [ 2.56 ] ] ) , assuming the small and large scale - dependence of the generating function factorize , for which we calculate the shape of the bias parameter and of the higher order statistical properties . in fig .
[ bx ] we present the function @xmath172 for the two models .
we also present the results for @xmath417 that give the statistical properties of the objects defined by a threshold .
the difference between the two cases is an indication of the magnitude of the variations that may occur within the hypothesis of hierarchical clustering . in both cases the bias is proportional to @xmath16 for large values of @xmath16 , and is weakly varying or nearly constant for small values of @xmath16 .
these results are natural consequences of the hierarchical clustering and by no mean have been imposed a priori .
the general properties of these quantities have been given in ( bes92 ) .
two properties of observational interest ( eqs .
[ [ 2.41a]-[2.41b ] ] ) are given in sect .
[ modindprop ] .
although we have not made any attempt here to transform the scaling parameter @xmath16 which caracterises the halos into luminosity , it is very obvious from both the fig .
5a of bes92 and fig . 5 of benoist et al .
( 1996 ) that , within the uncertainty in defining the luminosity , we can bring our predictions of the bias is in agreement with the data ( but it is also obvious from these figures that the process looks somewhat easier for the factorized model , e.g. fig .
4.7 of bernardeau 1992 , phd thesis ) . numerical simulations ( munshi et al . 1999 ) seem to favor the minimal model .
the new results that have been derived in this paper concern the high order correlation functions .
one can obviously calculate order by order the shape of the 4 , 5-point correlation functions of the halos .
it requires the expansion of the generating function @xmath228 with @xmath231 ( that can be done at the level of the equations [ [ 2.47a]-[2.47d ] ] ) .
the coefficients of this expansion , @xmath418 , @xmath419 , @xmath420 give the values of @xmath421 ( eq . [ [ 2.39b ] ] ) for growing values of @xmath422 : @xmath423 following a tree shape construction ( e.g. bernardeau 1992 ) .
the resulting values of @xmath339-@xmath424 are presented on figs .
[ s3]-[s6 ] .
they have been calculated for the minimal and the factorized models using @xmath425 .
one can see that the values of @xmath333 for the factorized model are always positive ( solid lines ) and above the positivity constraints .
this is not the case for the minimal - tree models ( dashed lines ) .
an alternative way to directly constrain the high order correlation functions of the halos is to observe their void probability function .
this is the probability @xmath426 that a spherical volume @xmath123 does not contain any object of a given kind , and it is closely related to the behavior of the high - order correlation functions of the objects ( white 1979 ) . in case of the tree - hierarchical models it is possible to relate @xmath426 to the generating function of the vertices @xmath427 ( bes92 -see also schaeffer , 1984- and appendix b ) . for the objects of kind @xmath16
we then obtain @xmath428 where @xmath429 is the value taken by @xmath430 depending on the values of @xmath16 and @xmath123 .
the calculable properties of @xmath426 are then direct consequences of the properties of @xmath431 . in fig .
[ sigma ] we present the shape of the function @xmath432 as a function of @xmath429 for various values of @xmath16 and for the two explicit models we considered . a change between the void probability function in the matter field ( assumed to be represented by a fair sample of points ) , and the void probability function of halos
this is due to the existence of biases in the high order correlation functions , that lead to changes in the values of @xmath333 from the matter field to the halo field , and are independent of the fact that @xmath433 or not .
for the two models there is a common limit for @xmath434 , corresponding to @xmath435 , which is independent of @xmath16 .
in such a case the void probability function reads , @xmath436 the resulting form for @xmath437 corresponds to the model of schaeffer ( 1984 ) with @xmath438 .
the dependence of @xmath437 with @xmath16 is quite weak at this level of description and most of the bias effects are contain in @xmath172 ( especially for the bright end ) .
the weak dependence of @xmath421 with @xmath16 is an illustration , at the level of the multi - point correlation functions , of this feature .
the kind of variation that is expected is , however , dependent of the form assumed for @xmath224 . for the form ( 2.55 )
this is a growing function of @xmath16 , for the form ( 2.56 ) a decaying .
we have thus no universal answer for a set of typical galaxies ( @xmath439 ) .
however , for rich clusters or bright galaxies we know what should be the shape of the void probability function .
the main advance brought by our models is that it can reconcile the measurements in catalogues and in numerical simulations in a unique description .
the properties of the field halos are mainly due to a threshold effect in the nonlinear density field ( specially for the rare halo limit ) , at contrast with the matter field properties which are solely due to the dynamics .
it is however quite remarkable to note that the `` model - independent '' requirements [ modindprop ] for the count - in - cells precisely call for the properties required ( [ rahalo ] ) for the above universal asymptotic behavior to hold .
the hamilton model , for instance , that does not satisfy these requirements , also does not lead to the above behavior . in this sense , the properties of the field halos definitely reflect the underlying matter dynamics .
simply , the conditions we imposed already for the count - in - cells to reflect that dynamics also induce the above behavior . in the previous section , correlation properties of the matter field , and correlation properties of the halos were considered separately .
but in fact the general results we get hold to describe the auto - correlation functions for a given kind of objects , as well as for the cross - correlations between objects of different kinds .
we present the results that follow as an illustration of what can be derived from this theory .
the calculation of the mass multiplicity function of the halos , representing galaxies , being in a cluster involves the statistics of a two component field .
one component is the matter field characterized by a density @xmath440 , a two - point correlation function @xmath441 and @xmath6-body correlation functions following a tree - hierarchical form given by the vertex generating function @xmath427 ; the second component is a halo field characterized by a parameter @xmath16 which leads to a density @xmath442 , and correlation functions given by @xmath228 .
as the latter function describes as well the auto - correlations as the cross - correlations between the density field and the halo field , one can compute the joint density distribution of the two fields , @xmath443 , which is the probability that the volume @xmath123 contains @xmath272 points of matter and @xmath444 objects of scaling parameter @xmath16 .
the generating function of such joint probability distribution is defined by @xmath445 and its expression reads , @xmath446 we are interested in the expectation number of objects of kind @xmath16 being in a volume @xmath123 that contains a certain mass @xmath447 , @xmath448 the mass @xmath449 is chosen so that the volume @xmath123 is rich enough to represent the inner part of a cluster ( @xmath450 ) . the calculation of @xmath451 from ( f.2 )
is quite straightforward .
the generating function of @xmath452 is simply @xmath453 whereas the one of @xmath454 is @xmath455 the latter expression can be deduced from ( f.2 ) , @xmath456 when we are the regime @xmath450 , we can use the continuous variable @xmath457 to describe the content of the volume @xmath123 instead of the variable @xmath272 .
we are thus led to calculate , @xmath458 when @xmath278 is large one can compute more precisely the previous relation .
its result is dominated by the singularity that lie in @xmath459 in the complex plane .
this singularity is characterized by the equations , @xmath460 the integrations in ( f.6 ) are then performed around the singularity @xmath461 which requires the expansion of @xmath427 and of @xmath228 around @xmath462 .
we eventually get , @xmath463 the mass distribution function in rich clusters @xmath464 normalized so that @xmath465 is then given by , @xmath466 that reduces to , @xmath467 we can notice that the result is independent of @xmath468 , say the richness of the cluster . for any tree - hierarchical model
the behavior of @xmath228 has been shown to be close to @xmath469 $ ] with @xmath172 proportional to @xmath16 ( eq . [ [ 2.53 ] ] ) , so that @xmath464 behaves roughly as @xmath470 at least when @xmath16 is large , @xmath471 as a result the upper cut - off of @xmath87 , characterized by @xmath94 ( eq . [ [ 2.19 ] ] ) , is shifted towards greater values of @xmath16 .
this change can be calculated as soon as a particular form for @xmath224 has been chosen . for the factorized - tree model it is possible to give an explicit result .
in this case , from eq .
( [ e.48 ] ) we have in the rich cluster limit , @xmath472,\ ] ] which , from eq .
( [ e.51 ] ) , can be written , @xmath473 this is the function which is plotted on fig . [ hc ] with a solid line .
thus , for the factorized - tree model we get the position of the cut - off for @xmath464 , @xmath474 , @xmath475 in case of the minimal tree - hierarchical model ( eq . [ [ 2.55 ] ] ) , the cut - off is given by , @xmath476 one can notice that the effect is similar , although somewhat more important .
this has important consequences . for a given cluster mass , the galaxy multiplicity function within the cluster
is expected to be biased towards higher masses , irrespective of other phenomena that can change the observed luminosity function . _
the @xmath477 ratio within clusters is thus expected to be biased and smaller than the average in the universe_.
in this paper we address the issue of the halo correlation properties in a strongly non - linear cosmic density field .
this is a dramatic extension of the results obtained by bardeen et al .
( 1986 ) and mo et al .
( 1997 ) for gaussian fields .
the difficulty is to have a reliable description of the matter density fields and its high order correlation properties .
our analyses are all based on the hierarchical properties ( in the sense of eq .
[ [ 1.1 ] ] ) which seem well established in numerical simulations or in the observable universe .
we further assume a tree decomposition of the correlation functions(see szapudi & szalay 1997 , 1998 for another approach ) .
these assumptions lead to two crucial properties : * factorazibility of the mass dependence of the correlation properties ; * the bias factors and vertices all depend only on the internal properties of the correlated halos , through a unique scaling variable @xmath16 . the central formula we derived is given by eqs .
( [ 2.47a]-[2.47d ] ) which explicit the generating function of the vertices that describe the halo auto - correlation function as well as halo - matter cross - correlations .
it provides adapted mathematical techniques and tools to explore the relation between the intrinsic and joint properties of the halos in a nonlinear density field .
an illustration of the strength of these methods is proposed in the last section .
generically , we find that in all relevant cases @xmath435 ( which corresponds to vertices that are all equal to unity ) in the rare halo limit .
also , the mass weighted statistics of the halo distribution reduces , when all dense halos are included , to the statistics of the matter field , as a generalisation of the properties known ( bes92 , schaeffer 1987 ) for the two - body correlation function .
this is an indication that the halo correlation properties , especially for the higher masses , reflect the selection effects more than the intrinsic properties of the matter field ( this behavior is similar to what is expected in gaussian field , e.g. mo et al . 1997 ) . in the small mass limit
the behavior of the correlation functions are more dependent on the underlying matter field .
robust results in all regime therefore rely on precise assumptions on the geometrical dependence of the correlation functions .
the simplest assumption is to suppose that the @xmath6-point functions are exact trees in the sense that the vertices are pure numbers .
this corresponds to what we call the minimal tree - hierarchical model . in this case
it is possible to relate the correlation properties of halos , and in particular to the @xmath16 dependence of @xmath172 , to their mass function .
it leads however to mathematical inconsistencies ( violation of positivity constraints ) that we were not able to cure within the minimal model . as an alternative
, we develop a purely phenomenological description , the factorized tree - hierarchical model .
we use this model for illustration purposes , and because in this case no mathematical inconsistencies are encountered .
our results for the bias are consistent with the observed luminosity dependence of the bias ( see discussion in bes92 and in benoist et al 1996 ) although we made no attempt here to determine the luminosity of the halos we have considered .
our results can also be checked against numerical simulations .
early results by bernardeau ( 1996 ) were focused on the quasilinear regime but more recently munshi et al .
( 1999 ) checked in some details the properties of @xmath417 in the light of the predictions that were made in our previous paper ( bes92 ) for the minimal model .
they found that the qualitative and quantitative behavior of @xmath417 were in perfect agreement to what we expected for the minimal tree model , which make the mathematical inconsistencies we found all the more puzzling .
obviously , further numerical investigations would be precious for constraining the models .
alimi , j.m . , blanchard , a. , schaeffer r. , 1990 , apj 300 , l35 balian , r. & schaeffer , r. 1988 , apj , 335 , l43 balian , r. & schaeffer , r. 1989 , a&a , 220 , 1 * ( bas89 ) * bahcall , n.a . , 1979 , apj , 232 , 689 bardeen , j. m. , bond , j. r. , kaiser , n. & szalay , a. s. 1986 , apj , 304 , 15 benoist , c. , maurogordato , s. , da costa , l. n. , cappi , a. , schaeffer , r. , 1996 , apj , 472 , 452 benoist , c. , cappi , a. , da costa , l. n. , maurogordato , s. , bouchet , f.r . , schaeffer , r. , 1998 , apj , in print bernardeau , f. 1992 , apj , 192 , 1 bernardeau , f. 1992 , thse de doctorat de luniversit paris 6 bernardeau , f. 1994 , a&a , 291 , 697 bernardeau , f. 1996 , a&a , 312 , 11 bernardeau , f. & schaeffer , r. 1991 , a&a , 250 , 23 * ( bes91 ) * bernardeau , f. & schaeffer , r. 1992 , a&a , 255 , 1 * ( bes92 ) * bouchet , f. , schaeffer , r. & davis , m. 1991b , apj , 383 , 19 colombi , s. , bouchet , f. r. , schaeffer , r. , 1994 , a&a , 281 , 301 colombi , s. , bouchet , f. r. , schaeffer , r. , 1995 , apjs , 96 , 401 colombi , s. , bouchet , f. r. , hernquist , l. , 1996 , apj , 465 , 14 colombi , s. , bernardeau , f. , bouchet , f. r. , hernquist , l. , 1997 , mnras , 287 , 241 davis , m. , efstathiou , g. , frenk , c. & white , s.d.m .
1985 , apj , 292 , 371 davis , m. , efstathiou , g. , frenk , c. & white , s.d.m .
1992 , nature , 356 , 489 davis , m. & peebles , p.j.e .
1977 , apjs , 34 , 425 fry , j.n .
1984a , apj , 277 , l5 fry , j.n .
1984b , apj , 279 , 499 fry , j.n .
1986 , apj , 306 , 358 groth , e. & peebles , p.j.e .
, 1977 , apj , 217 , 385 hamilton , a.j.s . ,
1988a , apj , 331 , l59 hamilton , a.j.s .
, 1988b , apj , 332 , 67 hamilton , a.j.s . ,
kumar , p. , lu , e. & mattews , a. 1991 , apj , 274 , 1 jannink , g. & des cloiseaux , j. 1987 , _ les polymres en solution , _ les ditions de physique , les ulis , france .
kaiser , n. 1984 , apj , 284 , l9 maurogordato , s. , schaeffer , r. & da costa , l.n .
1992 , apj , 320 , 13 mo , h. j. , jing , y. p. , white , s. d. m. , 1997 , mnras , 284 , 189 munshi , d. , bernardeau , f. , melott , a. , schaeffer , r. , 1998a , astro - ph/9707009 munshi , d. , coles , p. , melott , a. , 1998b , astro - ph/9812337 munshi , d. , coles , p. , melott , a. , 1999 , astro - ph/9902215 peebles , p.j.e .
1980 , the large scale structure of the universe , princeton university press , princeton , n.j .
, usa peebles , p.j.e . , 1987 , _ large - scale motions in the universe _ , eds v.c .
rubin , g.v .
coyne , princeton , princeton university press saslaw , w.c .
& hamilton , a.j.s .
1984 , apj , 276 , 13 saunders , w. et al .
1991 , nature , 349 , 32 schaeffer , r. 1984 , a&a letters , 134 , l15 schaeffer , r. 1985 , a&a letters , 144 , l1 schaeffer , r. 1987 , a&a letters , 180 , l5 schaeffer , r. & silk , j. 1985 , apj , 292 , 319 schechter , p. , 1976 , 203 ,
297 scoccimarro , r. , colombi s. , fry , j. n. , frieman j. a. , hivon e. , melott , a. , 1998 , apj , 496 , 586 scoccimarro , r. , frieman j. a. , astro - ph/9811184 smoot , g.f . ,
1992 , apjl , 396l , 1 szapudi , i. szalay , a. , 1997 , apj , 481 , 1 szapudi , i. szalay , a. , 1998 , apj , 494l , 41 valageas , p. , schaeffer , r. , 1997 , a&a , 328 , 435 * ( vs97 ) * valageas , p. , schaeffer , r. , 1998 , astro - ph/9812213 valageas , p. , lacey , c. & schaeffer , r. , 1999 , astro - ph/9902320 valls - gabaud d. , alimi , j.m . ,
blanchard a. , 1989 , nat . , 341 , 215 white , s.d.m .
1979 , mnras , 186 , 145
let us divide the universe into cells of volume @xmath7 .
the probability to find @xmath272 particles of matter in a given cell , @xmath116 , can be evaluated by means of the generating function , @xmath478 with , @xmath479 where the integral is calculated in the complex plane around @xmath480 .
the generating function @xmath481 is related ( bas89 , white 1979 ) to the integrals of the generating functions @xmath482 for the moment , we assume that the @xmath6-body correlation functions take the form ( 2.5 ) of the minimal tree - hierarchical model , @xmath483 where @xmath31 is a particular tree topology connecting @xmath29 points without making any closed loop , @xmath32 is a particular labeling of the point coordinates in a given topology @xmath31 , the last product is made over the @xmath29 links between the @xmath6 points ( labeled by their coordinates @xmath484 ) , and @xmath36 is a normalization parameter associated with the order of the correlation and the topology @xmath31 involved that takes the form , @xmath485 the product is made over all the vertices of the tree , and is defined once the topology @xmath31 is specified , @xmath47 is the weight for each vertex of the tree and depends on the number @xmath34 of outgoing lines ( see an example of graph in fig .
as shown in detail in a previous paper ( appendix a of bes92 ) , the function @xmath486 can be constructed by the classical `` tree - graph '' summation jannink and des cloiseaux 1987 ) ( , jannink and des cloiseaux 1987 ) .
let us define , @xmath487 then @xmath486 is given by , @xmath488(\la-1)\,n\,\d^3\vr\nonumber\\ & & \hmove -{1\over2}\int_v(\la-1)\,n\,\d^3\vr \,\tau(\vr){\d\zeta\over\d\tau}[\tau(\vr)]\eqno(a.7)\\ \xib_p(v)&=&s_p\left[\xib_2(v)\right]^{p-1},\end{aligned}\ ] ] with , @xmath489 .
\eqno(a.8)\ ] ] the function @xmath490 $ ] represents the vertices at the point @xmath218 with an @xmath218-dependent weight @xmath491 for each outgoing line .
the latter represents the sum of all tree graphs starting at the position @xmath218 .
its value is given by the implicit equation ( [ a.8 ] ) which states that @xmath491 equals the specific contribution of the first line @xmath492 multiplied by any kind of vertex @xmath47 , the @xmath153 remaining lines being dressed by as many @xmath493 factors .
the sum over all possible vertices introduces the function @xmath494 $ ] ( and not the function @xmath495 $ ] since one line is treated apart ) and an integration over @xmath496 is required to finally construct @xmath491 . the fig .
[ figa1]a presents a diagrammatic representation of the implicit equation ( a.8 ) .
an important remark that can already be made at this stage is that , although the matter particles outside the cell are strongly correlated with the ones inside , _ the generating function for the counts in a cell depends only on the correlations within this cell .
_ the next step is to simplify considerably the equations ( [ a.7]-[a.8 ] ) by means of a `` mean field '' approximation that turned out to be extremely accurate ( about @xmath497 to @xmath498 , see discussion in bes92 ) .
we integrate ( a.8 ) over @xmath218 and then approximate @xmath491 by a constant @xmath67 . the equation ( a.8 ) is then changed in , @xmath499 with @xmath500 the equation ( a.7 ) then reads , @xmath501 this is the form that is used for practical calculations . in the limit of a continuous distribution ,
when @xmath440 goes to infinity for a given mass density @xmath502 , the product @xmath503 remains finite .
it will turn out to be convenient to define in this case the variable @xmath504 which remains finite .
we are then led to use the function @xmath59 , @xmath505 and the relations ( [ a.9]-[a.11 ] ) then read , @xmath506 so far we considered the minimal tree - hierarchical model , for which the coefficient @xmath36 are given by ( a.5 ) . in case of a general tree - hierarchical model , the equations ( [ a.6]-[a.8 ] ) have to be modified to take into account the dependences of the coefficients @xmath36 with the geometry of the vertices , @xmath507 where the product is still over all the vertices associated with the tree topology @xmath31 , @xmath50 is the coordinate of the considered vertex and @xmath51 are the coordinates of the end points of the @xmath34 outgoing lines of the vertex @xmath440 .
the difference with the previous calculation is that the vertex now depends on the geometry of the outgoing lines . in this case
we are led to define the function @xmath508 representing the sum of all tree graphs starting with the line that joins @xmath218 and @xmath496 .
the function @xmath491 defined previously is simply the integral over @xmath496 of @xmath508 within the volume @xmath7 .
the analogous of the equation ( a.8 ) now reads ( see also fig .
[ figa1]b ) @xmath509.\nonumber\end{aligned}\ ] ] it can be easily checked that this equation gives ( a.8 ) when the vertices are pure numbers by an integration over @xmath496 . the function @xmath486 is now given by ( @xmath510 ) , @xmath511 which also reduces to ( a.7 ) in case of the minimal model . in the general case
there is still a possible `` mean field '' approximation which just assumes that @xmath508 is proportional to @xmath492 , @xmath512 we are then led to exactly the same equation ( [ a.9]-[a.14 ] ) for @xmath486 and @xmath59 with @xmath70 given by ( a.6 ) and the vertices @xmath47 defined by the expectation values @xmath513 obviously , the validity of such an approximation depends on the strength of the geometrical dependences of @xmath514 .
we assume here that they are weak enough so that the mean field approximation is valid .
in the limit of a continuous distribution of matter we have the relations @xmath515 with , @xmath516 where @xmath517 is the mean mass contained in a cell of volume @xmath7 .
the counts in cells are then obtained by a simple poisson convolution of eq .
( [ a.21 ] ) .
realistic models for @xmath116 and hence @xmath59 or @xmath70 have been widely discussed ( schaeffer 1985 , bas89 , bes92 ) .
generically @xmath59 must have singularities at negative @xmath68 and behave as @xmath518 at large @xmath68 , with @xmath519 .
in the strongly non - linear regime @xmath520 , ( [ a.21 ] ) can be simplified into , @xmath521 this form holds provided @xmath522 is small . for low values of @xmath1 , however , @xmath523 gets large and , @xmath524 may not be small enough .
the form ( [ a.30 ] ) thus holds only provided , @xmath525 the halos referred to in this paper are those for which ( [ a.31 ] ) is satisfied .
we use the same partition of the universe into cells of volume @xmath127 ( which are not necessarily the same ) as in the appendix [ aappend ] and consider a set of @xmath526 cells denoted by the index @xmath527 .
we aim to calculate the probability that @xmath528 elementary objects are in cell 1 at position @xmath154 , @xmath529 in cell 2 at position @xmath152 , ... the corresponding generating function is , @xmath530 in case @xmath531 is known , @xmath532 can be obtained through , @xmath533 we may ask for the probability that cells @xmath534 , which are assumed to all lie with a larger finite volume @xmath123 , contain respectively @xmath535 elementary particles , whatever the contents of the cells @xmath536 to @xmath526 are . this amounts to sum over @xmath537 in ( b.1 ) with @xmath538 all equal to unity .
it is important to note that in this case all integrations over the variables within cells @xmath536 to @xmath526 drop out : the sums simply run from @xmath539 to @xmath124 . in case
@xmath540 we simply recover the expressions ( [ a.1 ] , [ a.2 ] ) for the counts in one cell , the solution of which is given by the equations ( [ a.9]-[a.14 ] ) . from now on
we will always implicitly assume that the cells @xmath534 all lie within a volume @xmath123 , much larger than the volume @xmath127 ( @xmath541 ) of the elementary cells .
once again the minimal tree - hierarchical model is simpler and deserves to be treated separately .
the function @xmath542 can be constructed in a similar way than in ( [ a.7 ] , [ a.8 ] ) .
the only change is that the value of @xmath543 depends on the volume in which the corresponding point lies .
the equation for @xmath491 now reads , @xmath544 .
\eqno(b.3)\ ] ] and the expression of @xmath542 is , @xmath545\,\nonumber\\ & & \hmove \ \ \ \disp{-{1\over2}\sum_{i=1}^{\p}\ , \int_{v_i}(\la_i-1)\,n\,\d^3\vr\ , \tau(\vr){\d\zeta\over\d\tau}[\tau(\vr)]}. \eqno(b.4)\end{aligned}\ ] ] once again we can introduce a mean field approximation .
we assume that @xmath491 has a constant value @xmath546 within each volume @xmath125 .
the equation for @xmath546 can be obtained from ( b.3 ) , @xmath547 where @xmath548 is the mean value of @xmath492 within the volume @xmath125 and @xmath549 is its means value when @xmath218 is in @xmath127 and @xmath496 in @xmath550 .
for convenience , we split @xmath546 into two terms @xmath551 and @xmath552 , @xmath553 the notation @xmath551 corresponds to graphs the first line of which starts and ends within the same box @xmath125 , whereas @xmath552 corresponds to graphs that start in the box @xmath125 but immediately join another box @xmath554 .
the function @xmath555 is now given by , @xmath556 the same summation can be written in a different way that will turn out to be useful in the following .
we can first focus on one cell .
let us define @xmath557 by , @xmath558\nonumber\\ & & \hmove \ \ \ -{1\over2}(\la-1)\,n\,v\,\tau(\la,\theta){\d\zeta\over\d\tau } [ \tau(\la,\theta)+\theta ] , \eqno(b.9)\end{aligned}\ ] ] with , @xmath559 these relations define @xmath557 as a function of two variables @xmath543 and @xmath231 ( and the volume of the cell @xmath7 ) .
this function is the sum off all connected tree graphs within the cell ( with vertices only in the cell , all being connected by lines ) having an arbitrary number of outgoing lines , each weighted by a factor @xmath231 .
the generating function @xmath555 can be built up with the functions @xmath560 once @xmath231 is specified for each values of @xmath125 .
the basic property required for such a construction is that one can relate the partial derivative of @xmath557 relatively to @xmath231 ( for a fixed @xmath543 ) to the derivative of @xmath74 , @xmath561 with @xmath562 given by ( b.10 ) . as a result the equation ( b.7 ) can be written using the function @xmath557 instead of the function @xmath74 , @xmath563 and the function @xmath555 now reads @xmath564 the equations ( b.12 ) and ( b.13 ) are similar to the equations ( [ a.9]-[a.14 ] ) , where @xmath70 has been replaced by @xmath557 as a function of @xmath231 and where the matter have been replaced by cells of weight @xmath565 .
the function @xmath557 is the sum of all diagrams within a given cell when the environment of the cell has a weight @xmath231 .
the function @xmath486 in the appendix a corresponds to the case when no constraints have been put in the vicinity of the cell @xmath125 , that is given by @xmath303 , so that @xmath566 .
the form @xmath557 is the central relation for the construction of the statistics of the content of a large cell @xmath123 in terms of the elementary finite cells @xmath127 , instead of the elementary points of matter as in the appendix [ aappend ] .
that will be the ground for joint count - in - cell calculations .
the case of the general tree - hierarchical models is basically the same although technically more complicated .
the equation for @xmath508 now reads when @xmath496 is in the cell @xmath567 , @xmath568 .
\eqno(b.14)\end{aligned}\ ] ] the expression of @xmath542 has to be changed in a similar way , @xmath569 the extension of the mean field approximation leads to assume that , @xmath570 where @xmath218 is in the volume @xmath125 and @xmath496 in the volume @xmath567 .
the value of @xmath551 corresponding to the definition ( b.6 ) is , @xmath571 whereas , @xmath572 our aim is to calculate ( b.15 ) with the mean field approximation .
the quantities @xmath573 in fact depend not only on @xmath574 but also on the positions of the cells in the volume @xmath123 . in case of the minimal tree - hierarchical model
this dependence is completely determined by the values of @xmath549 .
for the general case the vertices introduce some extra geometrical dependences .
we then are led to introduce a second mean field approximation which is to assume that @xmath573 is just proportional to @xmath549 as far as the positions of the cells are considered .
it just leads to change the equation ( b.16 ) in , @xmath575 when @xmath218 and @xmath496 are in @xmath127 and , @xmath576 when @xmath218 and @xmath496 are in two different cells @xmath127 and @xmath550 .
@xmath259 is the mean value of the two - point matter correlation function in the large volume @xmath123 .
the transformation of the relation ( b.14 ) and ( b.15 ) introduces the geometrical averages , @xmath577 the index @xmath34 is the total number of lines and @xmath178 is the number of long lines ( going from one cell to another ) .
as soon as @xmath578 , the averages @xmath221 are independent of the partition of @xmath123 in small cells .
these generalized vertices are thus specific of the matter correlation properties .
we then define the generating function @xmath224 of this vertices @xmath579 for the minimal tree - hierarchical model we simply have @xmath580 the equation for @xmath551 and @xmath552 now reads @xmath581 and @xmath582 the expression of @xmath555 can be transformed as previously in a tree - graph summation .
we define @xmath557 by @xmath583\nonumber\\ & & \hmove \ \ \ -{1\over2}(\la-1)\,n\,v\,\tau(\la,\theta ) { \partial\zeta\over\partial\tau}[\tau(\la,\theta),\theta ] \eqno(b.26)\end{aligned}\ ] ] with @xmath584 .
\eqno(b.27)\ ] ] we get an equation similar to ( b.11 ) , @xmath585 and the function @xmath555 is the same than in ( [ b.12 ] , [ b.13 ] ) : @xmath586 the only difference between the minimal tree model and the general case is that in the latter we are led , in order to get tractable calculations , to integrate out the position dependence of the @xmath124 cells , and thus replace @xmath549 in ( b.12 ) by @xmath259 .
the information we get from the spatial dependences of the halo correlation functions is then poorer in this case . nevertheless , for a very large variety of applications ( see schaeffer 1984 , 1985 , 1987 , and the extended discussion in bes92 ) we can accurately use the averages of these correlation functions in the volume @xmath123 , as it is shown in the next appendix . the equations ( b.13 ) and ( b.30 ) are decisive results for joint counts in cells calculations .
they deserve attention .
the function @xmath555 is the sum of all connected tree graphs with their proper weights , as explained in the beginning of this section .
these relations mean that such a summation can be performed in two steps .
indeed one can first compute the sum of all graphs having their vertices in a given cell and a given number , say @xmath178 , of outgoing lines ( fig [ figb1 ] ) .
let us denote @xmath587 such a sum .
the generating function for these graphs is precisely the function @xmath557 ( eqs . [ [ b.9 ] ] and [ [ b.26 ] ] ) calculated in the given cell , @xmath231 being the weight attached to each free line .
the expansion of @xmath557 in powers of @xmath231 , @xmath588 then leads to the corresponding sums @xmath587 .
once the sum within an elementary cell is known , @xmath555 is obtained by connecting ( fig .
[ figb2 ] ) the different cells in all possible ways , each connection between , say cell @xmath125 and @xmath567 , having a weight @xmath549 for the case of the minimal tree model , or simply @xmath259 for the general case .
these connections build a new tree whose elementary vertices are the graphs already summed within one cell . at this point
however loops between cells are not excluded .
they arise when a cell is reached again after having been connected to various neighboring cells .
it can be seen in fig .
[ figb3]a that the original graph , at the matter point level , has no loops according to the rules of the general hierarchical model although the graph at the cell level contains one loop . from the expression of the generating function @xmath555 of the connected graph we now seek for the probability @xmath589 using the general formula ( b.2 ) .
an explicit calculation is possible when one assumes that the distances between the cells are large compare to their size so that we have , @xmath590 let us first inspect the structure of @xmath555 with the assumption ( [ b.36 ] ) , having in mind we seek a perturbation expansion in @xmath591 .
the function @xmath555 is a sum of tree graphs ( fig.[figb3 ] ) connecting points within the various cell @xmath592 , each connection being weighted by the average @xmath258 of the correlation function within the cell , and points belonging to two different cells , say @xmath125 and @xmath567 , with a weight @xmath593 . that this is the case in general is readily seen from equation ( b.31 ) where the weight of an external line appears indeed to be proportional to @xmath593 whereas an internal line has the weight @xmath14 .
the condition ( b.36 ) insures that a graph that contains a loop ( that is coming back to a given cell after having visited other cells ) introduces a factor @xmath594 more than the same graph ( same number of points in each of the cells ) ( fig .
[ figb3]b ) that does not contain the loop . in the limit ( b.36 ) , at leadind order , the remaining contribution to @xmath555 is the sum with the smallest number of external lines connecting the different cells .
whence , in the limit ( b.36 ) , each cell appears only once in each remaining graphs .
each cell ( as in fig .
[ figb2 ] ) then appear as a dressed point ( see fig . [ figb1 ] ) with outgoing lines that connect this cell to other ones .
the sum of the remaining graphs is then a sum of product of factors @xmath595 corresponding to different cells : @xmath596 where @xmath597 is independent of the parameters @xmath543 .
this , however is true for @xmath555 whereas ( b.2 ) involves @xmath598 $ ] .
we therefore should extend the above examination to the latter .
while the former is the sum of all connected tree graphs , the latter is the sum of all , connected or disconnected , tree graphs where now the contribution of graphs in which all points are not linked together is also included ( the same way that the generating function of the cumulants is given by the logarithm of the generating function of the moments ) . from fig .
[ figb4]a it is clear the expansion of @xmath598 $ ] into tree graphs , not only brings in additional disconnected graphs but also , as previously , loops between boxes .
we now proceed by selecting the connected contributions at the cell level , that is by identifying those graphs in which all cells are connected together ( but with possible disconnected parts within the cells ) .
these contributions may thus be built from disconnected diagrams at the matter level , that become connected at the cell level .
however for the same reason as for fig .
[ figb3]a , diagrams similar to fig .
[ figb4]a that build up loops among the cells have a negligible contribution : it is always possible to change a @xmath593 factor by a @xmath14 factor and to preserve the connection between the boxes .
this is not the case for diagrams like in fig .
[ figb4]b : in such a case changing a @xmath593 factor by a @xmath14 factor breaks the connection between the cells .
it is therefore not possible a priori to neglect the latter contributions .
eventually we then are left with diagrams that form trees between cells in which the `` dressed '' vertices of order @xmath178 corresponding to each cell are obtained by the summation of all diagrams - connected or not within the cell- having @xmath178 outgoing lines .
we know that the generating function of the connected diagrams in a cell , with outgoing lines weighted by @xmath231 , is @xmath599 .
the generating function of all diagrams , connected or unconnected , is therefore @xmath600 $ ] .
this point can be illustrated for the diagrams with one external line ( e.g. on fig .
[ figb5 ] ) .
the connected diagrams with one external line are given by @xmath601 .
adding diagrams with no external lines amounts to multiply @xmath601 by the generating function of those diagrams , @xmath602 $ ] and the fist derivative of @xmath603 $ ] with respect to @xmath231 is indeed @xmath604 $ ] . in a similar way
than at the end of the appendix [ aappend ] , we may introduce the function @xmath240 defined by , @xmath605 that changes the equation ( [ b.26 ] , [ b.27 ] ) in @xmath606 @xmath68 being defined in ( a.12 ) .
then in case we are only interested in large values of @xmath272 , in such a way that , @xmath607 then @xmath557 is of the order of @xmath608 ( @xmath68 is finite in eq .
[ b.32 ] , see the discussion at the end of [ aappend ] for the case where ( b.36.1 ) is not fulfilled ) .
so , @xmath600\approx 1+\chi(\la,\theta)$ ] so that only connected diagrams contribute to the dressed vertices .
this is also the result we could have obtained directly by writing @xmath609
\sim 1 + \chi(\la_1,\ldots,\la_{\p})$ ] , as can be inferred from the discussion at the beginning of this paragraph . in all cases ,
whether @xmath272 is larger than @xmath610 or not , each cell appears only once in each term of the expansion of @xmath609 $ ] : at the cell level all diagrams are connected trees .
the integration ( b.2 ) over @xmath574 is thus easily performed , and amounts to replace each factor depending only on the variable @xmath611 of one specific cell by its integral @xmath612 .
this simply amounts to replace the sum in a given cell of all graphs with external lines @xmath600 $ ] by , @xmath613 and to use these new generating functions as elementary vertices . for dense cells
( with the meaning of eq .
[ [ b.36.1 ] ] ) this expression can be simplified as @xmath614 as for the matter distribution ( eq . [ [ a.6 ] ] ) , we can define the generating function of the vertices @xmath615 ( app . [ aappend ] ) by , @xmath616 recalling that @xmath617 .
the expansion of @xmath615 in powers of @xmath231 defines new vertices @xmath618 , @xmath619 that replace the parameter @xmath47 relevant for the matter distribution .
we have , for instance @xmath620\,e^{\chi(\la)}.\end{aligned}\ ] ] in the continuous limit we can further make the change of variable @xmath621 ( e.g. eq . [ a.13 ] ) and we have , @xmath622\ \nonumber\\ & & \times\ e^{-{\varphi(y)/\xib}+y\,{m/\mc } } , \label{nu2compl}\end{aligned}\ ] ] where @xmath159 and @xmath623 are successive derivatives of @xmath240 with respect to @xmath231 . for the dense cells we have considered in this paper , @xmath134 , that is typically @xmath624 , these vertices , as well as the ones of higher order depend on mass only through the ratio @xmath625 .
this limit will be examined in detail in append .
[ dappend ] . for the underdense cells , that can not be called `` halos '' , that is for @xmath626 and typically @xmath256 , in many cases
, there is another scaling law , similar to the one found for the count - in - cells ( bas89 ) as well as for the bias parameter @xmath197 ( bes92 ) .
the vertices , in this case depend on mass only through the parameter @xmath627 , with an overlapping regime for @xmath628 .
this case , its conditions of existence and its implications for the cell correlations , will be studied elsewhere .
the sum over tree connections in all possible ways among the cells under consideration is then performed by writing , @xmath629 \eqno(b.41)\end{aligned}\ ] ] with for the expression @xmath630 of this weight after integration over @xmath631 ( when the loops are not included ) as for the weight @xmath632 in equation ( b.30 ) , although the two quantities are not the same .
] , @xmath633 in case of the general tree - hierarchical model @xmath549 in ( b.42 ) has to be replaced by @xmath259 .
we may similarly ask for the correlation among cells which contain more than @xmath272 elementary particles , and explore the effect of a threshold in the matter field .
the same calculation shows that @xmath422 , and @xmath74 are respectively replaced by @xmath634 the expansion of @xmath635 in powers of @xmath231 defines the vertices @xmath636 that are to be used to construct the many - body correlation functions among cells containing more than @xmath272 particles .
similarly , we may also affect a multiplicity @xmath272 to each cell that contains @xmath272 particles .
the related statistics can also be described by tree correlations with the functions @xmath637 the vertex generating functions obey the normalization conditions , @xmath638 in the first sum , the underdense cells are heavily weighted , with a negligible contribution of order @xmath639 , for the same reasons as the ones discussed in sect . [ pn ] for the count - in - cells .
the last , on the other hand , is heavily weighted towards the dense cells , the underdense contribution vanishing as a power of @xmath639 .
these sum rules imply for the vertices @xmath618 , for @xmath640 ( @xmath641 by convention ) @xmath642 let us consider a volume @xmath123 , made of finite elementary cells of finite volume @xmath7 .
we assume that , @xmath643 in the spirit of the previous section where we calculated the correlation among dense cells , we consider now the counts of dense cells in the volume @xmath123 , that is the probability to find a given number of `` full '' sub - cells in the volume @xmath123 . to be precise we may decide a cell is `` full '' if it contains more than @xmath272 elementary particles ( or exactly @xmath272 .. ) and `` empty '' when the converse is true .
the cells in the volume @xmath123 may be labeled from @xmath539 to @xmath644 .
we assume the threshold value @xmath272 to be the same for all cells although a generalization with a mixture of thresholds is straightforward . the generating function for the counts of cells in volume @xmath123
, @xmath645 can be defined from the probability @xmath646 of having @xmath647 full cells in @xmath123 . from the probability @xmath648 for a cell to be full , we may define a number density of cells @xmath649 now , since we know that the many - body correlation functions , in the limit @xmath650 , are given by a tree - graph series with a vertex generating function @xmath635 , we get the equation @xmath651 which are the same as ( a.7 ) , ( a.8 ) , but written for finite cells and with a different vertex generating function @xmath635 that replaces @xmath70 of equation ( a.6 ) . for the general tree - hierarchical models @xmath549 has to be replaced by @xmath259 . in case
all the volumes of the elementary cells @xmath127 are equal , and all thresholds @xmath652 are the same , we get the analogue of equations ( [ a.9]-[a.14 ] ) : @xmath653 these equations together with the equations ( b.26 ) , ( b.27 ) and ( b.43 ) allows to derive the generating function of the counts of halos in the volume @xmath123 .
once @xmath654 is known , one can get , by the inversion of the relation ( b.49 ) , the values of @xmath655 .
in particular the void probability function in @xmath123 ( the absence of dense objects , not the absence of matter ) can be directly obtained by @xmath656 $ ] .
we can note that in case the threshold is given by @xmath657 we should recover the expression of the void probability function for the matter field in the volume @xmath123 which reads , @xmath658 to check this we have to show that @xmath659 . from ( b.43 ) we get @xmath660 the latter , from ( b.26 ) , ( b.27 ) is given by @xmath661 if @xmath231 is finite , then @xmath662 must be finite too , and this is only possible when @xmath68 is small ( since @xmath578 ) and whence @xmath67 is small too .
if @xmath231 is large , then according to the equation ( b.63 ) , @xmath67 is necessarily small whatever the value of @xmath68 : its value is dominated by the environment . when @xmath67 is small @xmath663 if the function @xmath224 is regular enough , otherwise that would mean the successive mean field approximations that we did are not valid ( because the dependence of the vertices is too strong ) .
except in these cases we see that we recover the right form for @xmath664 .
in this section we are interested in cases where dense cells only are considered .
it means that we assume that @xmath665 so that we can apply the formula ( [ b.38.b ] ) .
we further assume that the density of points is large enough so that the continuous limit can be used .
it amounts to make the change of variable from @xmath631 to @xmath68 .
one can then use the formulae ( [ b.32]-[b.34 ] ) for @xmath599 .
the continuous limit is obtained by letting the number density @xmath440 of points go to infinity , while keeping the mass density @xmath502 finite .
when this limit applies , the cell content is described by its mass @xmath1 ( see sect .
remaining in the dense cell regime , the results can be described with the variable @xmath416 , and we have , @xmath666 the function @xmath282 generates by a series expansion , @xmath667 the vertices @xmath668 that weight the vertices of the tree expansion of the cells .
the number density @xmath442 of @xmath16-type cells is in this case @xmath669 in case we define a `` full '' cell by a threshold @xmath16 , @xmath229 and @xmath87 have to be replaced by , @xmath670 and in case each cell is weighted by its content they have to be replaced by , @xmath671 the function @xmath282 , eq .
( d.5 ) , and the corresponding generating functions deduced from @xmath672 or @xmath673 defines a new set of many - body correlation functions . for instance , the vertex , @xmath674 or its analogues @xmath675 , @xmath676 are simply the bias factors , @xmath172 , @xmath417 , @xmath677 respectively as presented in bes92 . in case of a single set of objects , the observable quantities , such as the various moments of the density distribution , give information only on @xmath682 since the absolute strength of the matter correlation is not directly observable .
the exact form of @xmath228 depends on the form taken by @xmath224 .
there is however a limit which is independent of the exact form of @xmath224 which corresponds to the case @xmath684 .
first of all we derive the expression of the bias , @xmath685 and @xmath87 is given by the equation ( d.4 ) .
the behavior of @xmath172 for large values of @xmath16 can be calculated by the steepest descent method , taking advantage of the singularity in the implicit equation for @xmath122 .
we are then led to define @xmath79 and @xmath686 so that @xmath687 , @xmath688 then we can derive the expression of the bias , @xmath689 as a result we get the behavior of @xmath172 when @xmath16 is large , @xmath690 the form ( d.4 ) can be calculated with a similar method .
the position of the singularity has to be determined as a function of @xmath231 .
so we consider @xmath691 and @xmath692 so that @xmath693 : @xmath694 we obviously have @xmath695 and @xmath696 .
we can then write the form of the generating function of the vertices , @xmath697 the resulting expression for @xmath431 is given by @xmath698 and as @xmath172 is proportional to @xmath16 , in the limit @xmath684 we have @xmath699 .
it implies that @xmath700 which leads to @xmath701 in the limit @xmath434 . this is a generalization of previous results obtained by bes92 where it was shown that @xmath702 in such a limit .
here we obtain that all the vertices @xmath703 .
this result is valid for the minimal tree - hierarchical model as well as for the general case . in the latter case
however we can only give the mean behavior of the many - body correlation functions of the halos as a function of scale and the geometrical dependence is not known .
we thus rather have to define the averages of the halo correlation function in a volume @xmath123 , @xmath704 and the coefficient @xmath286 by @xmath705 these coefficient are mere combinations of the vertices @xmath264 , and in the limit @xmath434 we get that @xmath286 is simply the number of trees that can be constructed with @xmath6 points , that is @xmath706 the latter limit is valid for any tree - hierarchical models , provided some regularity conditions hold for the generating function @xmath224 so as the above expansions are justified . | the question we address in this paper is the determination of the correlation properties of the dark matter halos appearing in cosmic density fields once they underwent a strongly nonlinear evolution induced by gravitational dynamics .
a series of previous works have given indications of what non - gaussian features are induced by nonlinear evolution in term of the high - order correlation functions . assuming such patterns for the matter field , that is that the high - order correlation functions behave as products of two - body correlation functions , we derive the correlation properties of the halos , that are assumed to represent the correlation properties of galaxies or clusters .
the hierarchical pattern originally induced by gravity is shown to be conserved for the halos .
the strength of their correlations at any order varies , however , but is found to depend only on their internal properties , namely on the parameter @xmath0 where @xmath1 is the mass of the halo , @xmath2 its size and @xmath3 is the power law index of the two - body correlation function .
this internal parameter is seen to be close to the depth of the internal potential well of virialized objects .
we were able to derive the explicit form of the generating function of the moments of the halo counts probability distribution function .
in particular we show explicitely that , generically , @xmath4 in the rare halo limit .
various illustrations of our general results are presented . as a function of the properties of the underlying matter field
, we construct the count probabilities for halos and in particular discuss the halo void probability .
we evaluate the dependence of the halo mass function on the environment : within clusters , hierarchical clustering implies the higher masses are favored .
these properties solely arise from what is a natural bias ( , naturally induced by gravity ) between the observed objects and the unseen matter field , and how it manifests itself depending on which selection effects are imposed .
( # 1)[#1 ] ( # 1)([#1 ] ) # 1 # 1 # 1;#2;#3;#4 ; # 1 , _ # 2 _ , * # 3 * , # 4 .
# 1;#2;#3 ; # 1 , _ # 2 _ , # 3 .
# 1;#2 ; # 1 , _ # 2_. ii ps . |
astronomical images obtained form the ground suffer from serious degradation of resolution , because the light passes through a turbulent medium ( the atmosphere ) @xcite before reaching the detector .
a number of methods was developed to alleviate this phenomenon @xcite and multiple special - case solutions were implemented as well ( e.g. @xcite ) , but none of them is able to provide a perfect correction and restore the diffraction - limited ( dl ) image .
therefore , the best observatories , in terms of angular resolution , are the spaceborne ones , since they are limited only by the dl .
recently , there is an increasing interest in the attempts to overcome the dl boundary .
devices which can give such possibility are called the quantum telescopes ( qt ) .
first qts will probably work in the uv , optical and ir bands , mainly because of the speed , maturity and reliability of the detectors .
latest progress in adaptive optics ( ao ) , especially so called extreme ao , makes the use of qts realistic also in ground - based observatories @xcite . in this letter
the general idea of quantum telescopes is considered . in particular
we refer to the setup proposed by @xcite , since , to our knowledge , it is the only existing detailed description of a qt . in fig .
[ fig : qtscheme ] we propose an upgraded version of this setup . according to @xcite , each photon coming from the extended source triggers a signal by qnd detection and gets cloned @xcite . the coincidence detector controlled by the trigger
is turned on for a short period @xcite and registers the clones .
after that it is quickly turned off , so that it receives only a small fraction of spontaneous emission from the cloning medium and virtually no clones from other photons . if the source is too bright and emits too many photons per unit of time , a gray filter should be installed . as a result , a set of clones is produced and registered .
the centroid position of the clones cloud is used to add 1 adu . ] at the corresponding position of the high resolution output image .
the exposure time has to be much longer than in classic telescopes ( ct ) , since ( a ) in most cases a narrowband filter has to be applied and ( b ) the qnd detection efficiency is much below 100% @xcite .
below we present the results of our detailed simulations of the qt system , discuss the feasibility of building such a device and predict its expected performance . to our knowledge , this is the first paper describing the detailed simulations of a qt of any design , as this is a preliminary concept . in our simulations as input images we used parts of real images obtained by the hubble space telescope ( hst ) .
such images are optimal for the purpose of qt testing because this observatory is working at its dl . we cropped and decimated them to 200@xmath0200 pixels to speed up the computations .
for the comparison , we also simulated the process of digital image formation in the case of a ct ( for a review on high angular resolution imaging see @xcite ) using the same images ( fig .
[ fig : obscuredpupil]e ) .
we assumed a real telescope for which the pupil is obscured by a secondary mirror and its truss ( spider ) @xcite .
we assumed a similar size of a secondary mirror and truss as it is installed on the hst ( fig .
[ fig : obscuredpupil]a ) . in the case of a ct , we sampled the counts ( photons ) from the original image ( fig .
[ fig : obscuredpupil]d ) and distributed them according to the airy diffraction pattern decimated to 61x61 pixels ( fig .
[ fig : obscuredpupil]b ) . for the simulation of a qt
, each photon from the reference image was converted to @xmath1 cloned photons ( fig .
[ fig : obscuredpupil]c ) .
for @xmath1 we assumed the poissonian distribution .
the clones were arranged using the gaussian profile ( see @xcite fig . 2 therein for justification ; we assumed @xmath2 = 10 ) centered around the photon s position . to include the effects of spontaneous emission
, on the top of it we added counts distributed equally within the coincidence detector plane and governed by the poissonian distribution .
the mean noise level was tuned to achieve a given signal - to - noise ratio ( snr ; see exemplary simulated exposure in fig .
[ fig : drawing]b ) .
the assessment of the snr was based on the comparison of the number of cloned photons with the number of counts originating from the spontaneous emission within the circular aperture of 3@xmath2 radius ( i.e. within the aperture , for which nearly all the clones are received ) .
it follows the approach of snr derivation presented in @xcite . in the next stage of simulations of the qt image formation , we computed the centroid of clones employing the matched filtering approach .
as the cloned photons exhibit gaussian spatial distribution @xcite , in our calculations the image registered on the coincidence detector was first convolved with the gaussian ( @xmath2 = 10 ) and then the centroid was obtained from the position of the maximum value of such a filtered image ( fig .
[ fig : drawing]c ) .
@xmath3 = @xmath4(@xmath5,@xmath6 ) ) generate @xmath7 generate @xmath1 clones in the @xmath7 around ( @xmath5,@xmath6 ) compute @xmath8 paste @xmath8 to @xmath9(@xmath5,@xmath6 ) we ran the simulations for different numbers of clones , reaching also very high numbers ( up to the expected value of ~10k , see @xcite for justification ) .
the mean level of the poissonian noise of spontaneous emission was set so that snr was : 3/1 , 2/1 , 1 , 1/2 , 1/3 , 1/4 , 1/5 , 1/6 , 1/7 , 1/8 , 1/9 , 1/10 , 1/11 , 1/12 , 1/13 and 1/14 .
such a selection of snr includesa value of 1/7.3 which was assessed for qt in @xcite .
the quality of the simulated qt outcomes was assessed by two indicators : peak signal - to - noise ratio ( psnr ) and mean centroid error ( rms value ) . for 16-bit pixel representation
the psnr measure is defined as follows : @xmath10 @xmath11 where @xmath12 and @xmath13 denotes the intensity of pixel at ( @xmath14 ) in respectively qt simulated image and reference high - fidelity image .
the psnr and the mean centroid errors are depicted in fig . [ fig : surfs ] .
the place where both dependencies meet each other was retrieved and presented in fig .
[ fig : betterworse ] .
it shows the minimal requirements for the clones number and the snr level , which should be satisfied to achieve the resolution enhancement in qt .
this curve can be treated as a first guidance for selection of qt parameters . as seen from figs .
[ fig : surfs ] and [ fig : betterworse ] , assuming snr @xmath15 7.3 , in average @xmath1657 clones per 1 detected photon are necessary to produce an image sharper than ct .
for @xmath1710k clones the image is virtually ideally restored , even for the lowest values of snrs .
the noticeable improvement of the outcome with higher clones count is related to the generally better estimation of the centroid when using the mf approach .
for very small number of clones , the photons distribution is dominated by isolated photon - detection events and the mf output is strongly dependent on actual arrangement of registered clones .
in contrast , for higher numbers of clones , the gaussian becomes more uniform and therefore , the mf provides much more reliable estimations even if the snr remains the same . in fig .
[ fig : outputimages ] we show exemplary resulting qt images for various snrs and clones counts .
the improvement in the image quality is easily noticeable as both parameters increase .
below we summarize and discuss the feasibility of the qts as emerging from the most recent literature . a constant progress in the qnd increases the qt feasibility .
there remains a possibility that some other kind of discriminator of spontaneous emission will be found .
possibly qnd does not require to hold the photon in a resonant cavity for several hundreds microseconds , as it was claimed before @xcite . according to @xcite , in the case of the cavity - free qnd the photons momentum change is also negligible .
however , this setup is suspected to produce false detections ( < 5% ) and can not be miniaturized .
authors are presently working on a version easier to miniaturize ( keyu xia priv . com . ) .
we see no strong limitations for the size of the device if it could be placed e.g. at the ` cassegrain ` or ` coud ` plane optically conjugated to the pupil plane of the optical system near the focus ( fig .
[ fig : focus ] ) .
another technological issue concerning the qt is the spontaneous emission from the process of cloning ( for details see @xcite ) .
as we showed , in our model , for a sufficiently large number of clones this problem could be overcome .
however , in the general case of the qt some more sophisticated methods might be needed ( e.g. different geometrical setup or filters ) .
given the recent progress in cloning , it might also become possible to clone photons from a wider wavelength span and at a greater amount @xcite .
in conclusion , the general idea of the quantum telescopes seems feasible and in the near future it might be possible to construct a technology demonstrator . in this paper we presented the updated schematic `` toy model '' of a qt system and the first quantitative results of simulations of such an imager , aiming at predicting the conditions which would guarantee its optimal performance .
we found that it is generally more important to provide more clones than to reduce the noise background .
given the predicted snr = 7.3 , a satisfactory results are obtained from less than 60 clones . using 10k clones ,
the signal is almost perfectly restored for any noise level .
we encourage the interest and discussion on the idea of the qt , since even a small increase in the resolution might lead to a major breakthrough in astronomical imaging . * funding . *
marian smoluchowski research consortium matter energy future from know .
polish national science center grant number umo-2012/07/b
/ st9/04425 .
polish national science center , grant no .
umo-2013/11/n / st6/03051 .
poig.02.03.01 - 24 - 099/13 grant : geconii - upper silesian center for computational science and engineering .
v. m. tikhomirov , `` dissipation of energy in isotropic turbulence '' , _ selected works of a. n. kolmogorov _ ( springer science + business media , 1993 ) http://dx.doi.org/10.1007/978-94-011-3030-1_47 [ online version ] beckers , j. m. , `` adaptive optics for astronomy
principles , performance , and applications '' , annual review of astronomy and astrophysics * 31 , * 13 - 62 ( 1993 ) http://adsabs.harvard.edu/abs/1993ara%26a..31...13b [ online version ] tonry , j. ; burke , barry e. ; schechter , paul l , `` the orthogonal transfer ccd '' , publications of the astronomical society of the pacific * 109 , * 1154 - 1164 ( 1997 ) http://adsabs.harvard.edu/abs/1997pasp..109.1154t [ online version ] roggemann , michael c. ; welsh , byron m. ; fugate , robert q. , `` improving the resolution of ground - based telescopes '' , reviews of modern physics . * * 69,**issue 2 , 437 - 505 ( 1997 ) http://adsabs.harvard.edu/abs/1997rvmp...69..437r [ online version ] hecquet , j. ; coupinot , g. , `` a gain in resolution by the superposition of selected recentered short exposures '' , journal of optics * 16 , * 21 - 26 ( 1985 ) http://adsabs.harvard.edu/abs/1985jopt...16...21h [ online version ] buie , m. w. and grundy , w. m. and young , e. f. and young , l. a. and stern , s. a. , `` pluto and charon with the hubble space telescope .
ii . resolving changes on pluto s surface and a map for charon '' , the astronomical journal , * 139 , * , 1128 - 1143 ( 2010 ) http://adsabs.harvard.edu//abs/2010aj....139.1128b [ online version ] v. korkiakoski , c. vrinaud , `` extreme adaptive optics simulations for the european elt '' , frontiers in optics , optics & photonics technical digest ( 2009 ) https://www.osapublishing.org/abstract.cfm?uri=aopt-2009-aotud1 [ online version ] kellerer , a. , `` beating the diffraction limit in astronomy via quantum cloning '' , astronomy & astrophysics , * 561 , * i d .
( 2014 ) http://adsabs.harvard.edu/abs/2014a%26a...561a.118k [ online version ] xia , keyu ; johnsson , mattias ; knight , peter l. ; twamley , jason , `` cavity - free scheme for nondestructive detection of a single optical photon '' , phys .
116 , 023601 ( 2016 ) http://journals.aps.org/prl/abstract/10.1103/physrevlett.116.023601 [ online version ] mutchler , m. ; beckwith , s. v. w. ; bond , h. ; christian , c. ; frattare , l. ; hamilton , f. ; hamilton , m. ; levay , z. ; noll , k. ; royle , t. , `` hubble space telescope multi - color acs mosaic of m51 , the whirlpool galaxy '' , bulletin of the american astronomical society * 37 , * p.452 ( 2005 ) http://adsabs.harvard.edu/abs/2005aas...206.1307m [ online version ] guerlin , c. ; bernu , j. ; delglise , s. ; sayrin , c. ; gleyzes , s. ; kuhr , s. ; brune , m. ; raimond , j.- . ;
haroche , s. , `` progressive field - state collapse and quantum non - demolition photon counting '' , nature * 448 , * 889 - 893 ( 2007 ) http://adsabs.harvard.edu/abs/2007natur.448..889g [ online version ] lamas - linares , a. ; simon , c. ; howell , j. c. ; bouwmeester , d. , `` experimental quantum cloning of single photons '' , science * 296 , * 712 - 714 ( 2002 ) http://adsabs.harvard.edu/cgi-bin/bib_query?arxiv:quant-ph/0205149 [ online version ] barbieri , m. ; ferreyrol , f. ; blandino , r. ; tualle - brouri , r. ; grangier , ph . ,
`` nondeterministic noiseless amplification of optical signals : a review of recent experiments '' , laser physics letters * 8 , * issue 6 , article i d . 417 ( 2011 )
http://adsabs.harvard.edu/abs/2011laphl...8..417b [ online version ] ralph , t. c. and lund , a. p. , `` nondeterministic noiseless linear amplification of quantum systems '' , american institute of physics conference series * 1110 , * p. 155 - 160
( 2009 ) http://adsabs.harvard.edu/abs/2009aipc.1110..155r [ online version ] s. e. derenzo , woon - seng choong and w. w. moses , `` fundamental limits of scintillation detector timing precision '' , physics in medicine and biology * 59 , * 3261 - 3286 ( 2014 ) http://dx.doi.org/10.1088/0031-9155/59/13/3261 [ online version ] p. grangier , j. a. levenson and jean - philippe poizat , `` quantum non - demolition measurements in optics '' , nature * 396 , * 537 - 542 ( 1998 ) http://www.nature.com/nature/journal/v396/n6711/abs/396537a0.html [ online version ] harvey , j. e. and ftaclas , c. , `` diffraction effects of telescope secondary mirror spiders on various image - quality criteria '' , applied optics .
* 34 , * 6337 - 6349 ( 1995 ) http://adsabs.harvard.edu/abs/1995apopt..34.6337h [ online version ] kellerer , a. , `` beating the diffraction limit in astronomy via quantum cloning ( corrigendum ) '' , astronomy & astrophysics , * 582 , * i d . c3 ( 2015 ) http://www.aanda.org/articles/aa/abs/2015/10/aa22665e-13/aa22665e-13.html [ online version ] c. w. helstrom , d. w. fry , l. costrell and k. kandiah , `` statistical theory of signal detection '' , international series of monographs in electronics and instrumentation , vol .
9 , ( pergamon press , oxford ; new york ) , 2nd edition ( 1968 ) http://www.sciencedirect.com/science/book/9780080132655 [ online version ] woodward , p.m. , `` probability and information theory with applications to radar '' , ( norwood , ma : artech house ) , 2nd edition ( 1980 ) http://www.sciencedirect.com/science/book/9780080110066 [ online version ] de martini , f. ; sciarrino , f. , `` investigation on the quantum - to - classical transition by optical parametric amplification : generation and detection of multiphoton quantum superposition '' , optics communications * 337 , * 4452 ( 2015 ) http://www.sciencedirect.com/science/article/pii/s0030401814007640 [ online version ] nogues , g. ; rauschenbeutel , a. ; osnaghi , s. ; brune , m. ; raimond , j. m. ; haroche , s , `` seeing a single photon without destroying it '' , nature * 400 , * 239 - 242 ( 1999 ) http://adsabs.harvard.edu/abs/1999natur.400..239n [ online version ] | quantum telescope is a recent idea aimed at beating the diffraction limit of spaceborne telescopes and possibly also other distant target imaging systems .
there is no agreement yet on the best setup of such devices , but some configurations have been already proposed .
+ in this letter we characterize the predicted performance of quantum telescopes and their possible limitations .
our extensive simulations confirm that the presented model of such instruments is feasible and the device can provide considerable gains in the angular resolution of imaging in the uv , optical and infrared bands . we argue that it is generally possible to construct and manufacture such instruments using the latest or soon to be available technology .
we refer to the latest literature to discuss the feasibility of the proposed qt system design .
copyright : the optical society ( osa publishing ) 2016 . published in optics letters vol .
41 no . 6 ( 2016 ) + |
the classification of the universality classes of nonequilibrium phase transitions to absorbing states is still an open problem of statistical physics @xcite .
reaction - diffusion ( rd ) models exhibiting phase transitions to absorbing states bear with a particular interest since many other types of systems like surface growth , spin systems or stochastic cellular automata can be mapped on them . unfortunately there hasnt been experimental verification of such classes except the coagulating random walk : @xmath0 ( crw ) in one dimension @xcite .
this is mainly due to the fact that the most well known , robust directed percolation ( dp ) class @xcite is sensitive to disorder @xcite , which occurs in real systems naturally .
it would be very important to find some other nonequilibrium class , which proves to be less sensitive to disorder , hence would provide a candidate for experimental verification .
the study of disordered systems is a hot topic of current research of statistical physics @xcite .
a principal condition for the relevancy of disorder is the harris criterion @xcite set up for equilibrium systems and has been found to be valid in some nonequilibrium models .
according to this criterion the pure critical point is stable against disorder if the spatial correlation length critical exponent @xmath1 fulfills the inequality @xmath2 where @xmath3 is the spatial dimensionality .
however an exception is reported very recently @xcite for dp with temporal disorder .
note that for crw ( which is exhibits the same scaling behavior as the @xmath4 annihilating random walk ( arw ) in 1d ) this criterion predicts relevant spatial disorder , ( @xmath5 ) still experiment @xcite did not report measurable effect of randomness unless very strong disorder fractures the medium .
besides the robust dp an other well known universality class is the so called `` parity conserving '' ( pc ) class of 1d nonequilibrium transitions .
this was discovered in a one - dimensional stochastic cellular automata ( ca ) exhibiting @xmath6 symmetric absorbing states and domain walls following even offspringed branching and annihilating random walk : @xmath7 , @xmath8 ( barw2 ) @xcite .
later it was observed by numerical studies of other models @xcite and field theoretical studies @xcite confirmed the existence of a corresponding fixed point distinct from that of dp . for a review see @xcite .
this class is also called as directed ising , dp2 or generalized voter model class .
according to the harris criterion disorder should be relevant for the critical behavior of this class ( @xmath9@xcite ) .
in contrast to this a recent renormalization group ( rg ) study @xcite did not find a strong disorder fixed point like in case of dp .
the question naturally arises if barw2 is really insensitive to disorder or the rg method @xcite is not applicable for this case .
the principal aim of the present study is to answer this question .
additionally in the absorbing phase of the barw2 model the arw dynamics dominates , which has also been addressed in the studies @xcite .
the renormalization study of arw with spatial randomness in the reaction rates found marginal perturbations to the fixed point of the pure system @xcite . on the other hand an exact study of the infinite reaction rate arw with space - dependent hopping rates found non - universal power - law decay of the density of a - s below a critical temperature @xcite .
note that in @xcite the strong disorder is defined in such a way that it can not completely block reactions or diffusion of the reactants .
therefore the so called infinitely strong fixed point of @xcite does not correspond to the blocking case .
such blocking or complete dilution was studied in a 1d toy model of random quantum ferromagnetic ising model @xcite where continuously variable power laws were found at the phase transition point .
the effect of disconnected domains in the reactions of crw and arw has been investigated in @xcite .
this study reported stretched exponential decay in case of exponential domain size distributions and continuously changing density decay for blocks distributed in a power - law manner . in the 1d model
we investigate such complete blocking may also occur , hence we investigate this topological effect .
to study pc class transitions with disorder we have chosen a very simple stochastic cellular automaton ( sca ) the nekimca introduced in @xcite .
it is easy to show that the dual variables of spins ( @xmath10 ) the kinks ( @xmath11 ) exhibit barw2 dynamics via the synchronous spin - flip dynamics . in this sca parity conserving kink branching is also generated due to the synchronous spin update of neighboring sites without introducing an explicit spin - exchange reaction as in case of the nekim model @xcite .
the reactions are like the followings : * random walk : @xmath12 * annihilation : @xmath13 * branching : @xmath14 in the nekimca there are two independent parameters parametrized as @xmath15 in the computer the state of a single spin is represented by a 1 or 0 of a 32 or 64 bit word @xmath16 ( depending on the cpu type ) . hence
32 or 64 ca samples ( exhibiting different random initial conditions but the same quenched noise ) updated at once . the following bit - parallel algorithm was used for the update of states @xmath16 at site @xmath17 . a random number @xmath18 is selected with uniform distribution .
if @xmath19 a spin - flip , corresponding to random walk of the dual variable @xmath20 \land s(j)\ ] ] is written to all bits of @xmath21 . following this another uniformly distributed random number @xmath22
is chosen and if @xmath23 a spin - flip , corresponding to annihilation of the dual variables @xmath24 \ \ & \ [ s(j+1 ) \land \
s(j ) ] ) \ \land s(j)\ ] ] is performed . here
@xmath25 denotes the quenched random noise variable with uniform distribution @xmath26 @xmath27 and @xmath28 are the logical xor and and of computer words .
note that for strong disorder @xmath29 or @xmath30 may be less than @xmath31 at a site meaning a blockade for that reaction .
a single monte carlo step ( mcs ) consists of updating all @xmath32 sites with periodic boundary conditions and writing back @xmath33 for @xmath34 ( throughout the paper the time is measured by mcs ) .
the simulations were carried out on @xmath35 sized systems . with periodic boundary conditions .
the initial states were randomly half filled lattices , and the density of kinks is followed up to @xmath36 mcs .
we started the exploration of the @xmath37 phase diagram by determining the phase transition of the impurity free ( @xmath38 ) case with @xmath39 .
this was located at @xmath40 as a power - law decay of the kink density @xmath41 with exponent @xmath42 .
this is in good agreement with the pc class value @xcite .
next we introduced the quenched disorder and determined @xmath43 for different @xmath44 values .
table [ pct ] and fig.[deltac ] show the results .
.numerical results for the disordered pc class transition line .
[ pct ] [ cols="<,<,<,<,<,<",options="header " , ] the spreading exponent @xmath45 changes similarly along the freezing transition line . it is close to @xmath46 for @xmath47 and deviates it significantly for @xmath48 ( see figure [ spreadins ] ) .
the average number of kinks ( @xmath49 ) evolves to a small constant value as @xmath50 ( @xmath51 ) all along the freezing transition line
. this value changes continuously by increasing the disorder strength . for @xmath52 only a single kink survives ( @xmath53 ) , while for other values see table [ freezt ] . in the frozen ( active ) phase in the long time limit pairs within the same block annihilate and basically lonely kinks are wandering in confined regions of sizes @xmath54 ( local @xmath55 reactions are also permitted with a small probability ) .
therefore the density of frozen kinks depend on the density of blocks , hence on @xmath44 .
the concentration of blocks @xmath56 can be expressed with the average length @xmath57 of blocks as @xmath58 for the exponential block size distribution ( [ ldist ] ) the average block size is @xmath59 with the no - wall probability @xmath60 the kink density is @xmath61 which for small wall probability @xmath62 has the leading order singularity @xmath63 this means that in the frozen phase @xmath64 .
the simulations confirm this . for @xmath65 , where only diffusion traps are present the
@xmath66 indeed ( see table [ freezt ] and figure [ betafr ] ) . for @xmath67 where annihilation is also blocked the exponent @xmath68 decreases by increasing @xmath44 ( see table [ freezt ] ) .
we have also investigated the phase space for positive @xmath69 ( with @xmath70 ) where @xmath71 causes a direct freezing transition via diffusion traps in the absorbing phase .
the density decay simulations ( see figure [ posdel ] ) show small deviations from the square - root law of arw ( [ sqrtlaw ] ) ) , which can be interpreted as corrections to scaling .
we have investigated the phase diagram of the nekimca as the function of quenched disorder with large scale simulations .
we have chosen nekimca , a very simple sca , which is known to exhibit pc class transition of kinks to absorbing states .
we determined static and dynamic exponents ( @xmath72,@xmath68,@xmath73,@xmath45 ) at several points along the phase transition line and in the inactive phase .
the main consequence of this survey is that weak disorder does not change the scaling behavior of the quantities studied .
this supports the results of a recent real space rg study @xcite but contradicts the harris criterion
. a possible resolution of this contradiction could be the study of the distribution of the critical point coordinates over the ensemble of samples of size @xmath74 .
according to recent progresses in the finite size scaling theory of disordered systems , thermodynamic observable are not self - averaging at critical points when the disorder is relevant in the harris criterion sense @xcite .
this lack of self - averageness at criticality is directly related to the distribution of pseudo - critical temperatures @xmath75 over the ensemble of samples @xmath76 of size @xmath74 .
this has been shown numerically in case of a wetting transition of a polymer chain with quenched disorder @xcite .
an interesting further direction of research would be the study of those distributions for nonequilibrium systems like for nekimca . for weak disorder ( @xmath77 ) we did not see deviations from the asymptotic square root power - law ( [ sqrtlaw ] ) in the inactive phase .
however for stronger disorder continuously changing density decay exponent ( @xmath72 ) were observed .
this corroborates a former rg study @xcite for arw with site dependent reaction rates and an exact calculation for the infinite reaction rate arw with disordered diffusion traps @xcite .
very strong disorder in our model may introduce complete blocking of reactions or diffusion with exponential domain size distribution . if the system becomes segmented by diffusion walls a freezing transition to fluctuating states occur . in this case in odd parity blocks residual particles remain active .
large scale simulations suggest that this kind of disorder is marginal , the density decay slows down from the ( [ sqrtlaw ] ) power - law by a logarithmic correction or changes continuously if the annihilation is blockaded too . in the frozen phase the concentration of residual kinks increases logarithmically , with @xmath78
this is in agreement with the results of @xcite , who considered the arw model with complete blockades .
however we did not see crossover to stretched exponential decay reported in @xcite for long times , but rather the decay slows down in our model .
this is due to the fact that in our model in the frozen state not only blockades exist , but reaction and diffusion probabilities are randomized , which can introduce slower power - laws @xcite . in the region of the phase diagram where the disorder is so strong that the annihilation reaction becomes segmented we found continuously changing exponents along the disordered pc and the freezing transition lines both .
* acknowledgements : * + we thank f. igli and g. m. schtz for the useful comments .
support from hungarian research funds otka ( grant no .
t046129 ) is acknowledged .
the authors thank the access to the niifi cluster - grid , lcg - grid and to the supercomputer center of hungary . for references
see : j. marro and r. dickman , , cambridge university press , cambridge , 1999 . for a review see : h. hinrichsen , adv
. phys . * 49 * , 815 ( 2000 ) . for a recent review
see : g. dor , _ phase transition universality classes of classical , nonequilibrium systems _ , rev .
phys . * 76 * , 663 ( 2004 ) .
r. kroon , h. fleurent and r. sprik , phys .
rev e * 47 * , 2462 ( 1993 ) .
w. kinzel , in _ percolation structures and processes _ , ed .
g. deutscher , r. zallen , and j. adler , ann .
soc . * 5 * ( hilger , bristol , 1983 ) ; h. k. janssen , z. phys .
b * 42 * , 151 ( 1981 ) ; p. grassberger , z. phys .
b * 47 * , 365 ( 1982 ) ; p. grassberger , z. phys .
b * 47 * , 365 ( 1982 ) .
h. k. janssen , phys .
e * 55 * , 6253 ( 1997 ) .
r. cafiero , a. gabrielli and m. a. muoz , phys .
e * 57 * , 5060 ( 1998 ) .
j. hooyberghs , f. igli , carlo vanderzande , phys .
e * 69 * , 066140 ( 2004 ) .
t. vojta and m. y. lee , phys .
e * 72 * , 036126 ( 2005 ) .
for a recent review see : f. igli , c. monthus , physics reports * 412 * , 277 ( 2005 ) a. b. harris , j. phys .
c * 7 * , 1671 ( 1974 ) .
a. j. noest , phys .
lett . * 57 * , 90 ( 1986 ) ; a. j. noest , phys .
e * 38 * , 2715 ( 1988 ) ; i. jensen , j. phys .
a * 38 * , 1441 ( 2005 ) .
p. grassberger , f. krause and t. von der twer , j. phys .
a : math.gen . , * 17 * , l105 ( 1984 ) .
h. takayasu and a. yu .
tretyakov , phys . rev
. lett . *
68 * , 3060 , ( 1992 ) .
i. jensen , phys .
e. * 50 * , 3623 ( 1994 ) .
n. menyhrd , j. phys .
a * 27 * , 6139 ( 1994 ) , m. h. kim and h. park , phys .
lett . * 73 * , 2579 , ( 1994 ) .
d. zhong and d. ben - avraham , phys .
a * 209 * , 333 ( 1995 ) .
n. menyhrd and g. dor , j. phys .
a. * 28 * , 4505 ( 1995 ) .
k. e. bassler and d. a. browne , phys .
* 77 * , 4094 ( 1996 ) .
h. hinrichsen , phys .
* e 55 * , 219 ( 1997 ) .
d. zhong , d. ben - avraham and m. a. muoz , eur .
j b * 35 * , 505 ( 2003 ) .
g. dor and a. szolnoki , phys .
e * 71 * 066128 ( 2005 ) .
j. l. cardy and u. c. tuber , phys .
lett . * 77 * , 4780 ( 1996 ) ; j. l. cardy and u. c. tuber , j. stat . phys . * 90 * , 1 ( 1998 ) o. al hammal , h. chat , i. dornic , m. a. muoz , phys .
lett . * 95 * , 100601 ( 2005 ) .
n. menyhrd and g. dor , brazilian j. of physics , * 30 * , 113 ( 2000 ) .
g. m. schtz and k. mussawisade , phys .
e * 57 * , 2563 ( 1998 ) .
p. l. doussal and c. monthus , phys .
e * 60 * , 1212 ( 1999 ) .
d. s. fisher , physica a * 263 * , 222 ( 1999 ) . c. mandache and d. ben - avraham , j. chem .
phys . * 112 * , 7735 ( 2000 ) .
b. p. lee , j. phys .
a * 27 * , 2633 ( 1994 ) .
l. braunstein , h. o. martin , m. d. grynberg and h. e. roman , j. phys .
a * 25 * , l255 ( 1992 ) .
s. wiseman and e. domany , phys rev e * 52 * , 3469 ( 1995 ) .
a. aharony , a.b .
harris , phys rev lett * 77 * , 3700 ( 1996 ) .
f. pzmndi , r.t .
scalettar and g.t .
zimnyi , phys .
lett . * 79 * , 5130 ( 1997 ) . s. wiseman and e. domany , phys
. rev . lett . * 81 * , 22 ( 1998 ) ; phys rev e * 58 * , 2938 ( 1998 ) .
a. aharony , a.b .
harris and s. wiseman , phys .
81 * , 252 ( 1998 ) .
k. bernardet , f. pzmndi and g.g .
batrouni , phys .
lett . * 80 * , 4477 ( 2000 ) . c. monthus and t. garel , j. stat .
( 2005 ) p12011 . | a stochastic cellular automaton exhibiting parity conserving class transition has been investigated in the presence of quenched spatial disorder by large scale simulations .
numerical evidence has been found that weak disorder causes irrelevant perturbation for the universal behavior of the transition and the absorbing phase of this model .
this opens up the possibility for experimental observation of the critical behavior of a nonequilibrium phase transition to absorbing state .
for very strong disorder the model breaks up to blocks with exponential size distribution and continuously changing critical exponents are observed . for strong disorder
the randomly distributed diffusion walls introduce another transition within the inactive phase of the model , in which residual particles survive the extinction .
the critical dynamical behavior of this transition has been explored . |
in this paper , we develop a multi - scale theory for groups of diffeomorphisms in the context of image registration .
the setting of _ large deformation diffeomorphic matching _ has been introduced in seminal papers @xcite and this approach has been applied in the field of computational anatomy @xcite .
the initial problem deals with the diffeomorphic registration of two given biomedical images or shapes .
an important aspect of this model is the use of reproducing kernel hilbert spaces ( rkhs ) of vector valued functions to define the lie algebra of the group of diffeomorphisms .
although our context is very different from kernel based methods for machine learning @xcite , some analogies may be worth noticing .
we now present the model developed in @xcite .
let @xmath0 be a domain in @xmath1 .
an admissible reproducing kernel hilbert space @xmath2 of vector fields is a hilbert space of @xmath3 vector fields such that there exists a positive constant @xmath4 , s.t . for every @xmath5 the following inequality holds @xmath6 the reproducing kernel associated with the space @xmath2 is given by @xmath7 , where @xmath8 are the pointwise evaluation maps which are linear forms by assumption on the space @xmath2 .
the kernel completely defines the space @xmath2 .
the diffeomorphism group @xmath9 associated with the rkhs @xmath2 is given by @xmath10,h)\right\}$ ] , where @xmath11 is the flow of @xmath12 , _
i.e. _ @xmath13 the diffeomorphic matching problem is the minimization of the functional @xmath14 where @xmath15,h)$ ] and @xmath16 are objects of interest such as groups of points , measures , currents or images .
the action of the group @xmath9 on the objects space is denoted by @xmath17 , where @xmath18 is an element of the group and @xmath19 is an object .
the distance function @xmath20 that enforces the matching accuracy is usually taken to be the square of the norm if the objects live in a normed vector space , _
e.g. _ for images one would use the @xmath21 norm , @xmath22 .
this minimization problem enables us to match images via geodesics on the group @xmath9 , if we endow @xmath9 with the right - invariant metric obtained by translating the inner product @xmath23 on the lie algebra @xmath2 to the other tangent spaces .
more importantly by its action on the space of images , the right - invariant metric on the group induces a metric on the orbits of the image space and the final deformation is completely encoded in the so - called initial momentum @xcite .
this initial momentum has the same dimension as the image .
since it is an element of a linear space , statistics can also be done on it @xcite .
a rkhs corresponding to a gaussian kernel is commonly used in practice .
the choice of its standard deviation @xmath24 is an important problem and describes a trade - off between the smoothness of the deformation and the matching accuracy .
indeed , a large standard deviation produces very smooth deformations with a poor matching accuracy of the structures having a size smaller than @xmath24 . on the contrary
, a small standard deviation results in a good matching accuracy but the deformations may present undesirable very large jacobians @xcite .
it is therefore a natural step to introduce a mixture of gaussian kernels with different standard deviations . in @xcite ,
the authors show that such kernels outperform single gaussian kernels , when registering images containing features of interest at several scales simultaneously : they provide a good matching quality , while keeping the diffeomorphisms smooth .
naturally , this introduces more parameters in the algorithm , which need to be tuned .
practical insight about how to parametrize the scales and weights of multiple kernels are given in @xcite .
the idea of using a mixture of kernels for matching is directly connected to @xcite , where it is proven that there is an equivalence between the matching with a sum of two kernels and the matching via a semi - direct product of two groups .
the work on the metric underlying the lddmm methodology @xcite has also been followed up by @xcite , where the authors introduce the notion of a bundle of kernels and argue that this general framework can be used to deal with multi - scale lddmm . by passing , we prove that their approach reduces to the mixture of kernels .
we give a self - contained and simple proof of this result based on lagrange multipliers and we develop an extension of @xcite , first to a finite number of scales and then to a continuum of scales .
the paper is divided into three parts : the first part focuses on a finite number of scales while the second treats the case of a continuum of scales .
the last part of the paper is devoted to numerical simulations , where we show in particular the decomposition on the given scales of the optimized diffeomorphism .
for the sake of simplicity , we first treat the case of a finite set of admissible hilbert spaces @xmath25 for @xmath26 . denoting @xmath27 ,
the norm proposed in @xcite as well as in @xcite is defined by @xmath28 the following lemma is the main argument to prove the equivalence between the approaches of @xcite and @xcite .
this lemma is an old result that can be found in @xcite .
however , we present for the sake of completeness a simple proof based on the lagrange multiplier rule . moreover , if one wants to skip the technical details of the proof , the formal application of the lagrange multiplier rule gives the result immediately .
we outline the formal proof : if one has to minimize the norm defined in formula then one can introduce a lagrange multiplier @xmath29 and obtain a stationary point of the lagrangian @xmath30 which gives @xmath31 and @xmath32 hence , it gives a heuristic argument why the problem of optimizing at several scale simultaneously reduces to a mixture of the kernel .
[ reductionlemma ] the formula defines a scalar product on @xmath2 which makes @xmath2 a rkhs and its associated kernel is @xmath33 , where @xmath34 denotes the kernel of the space @xmath25 .
let @xmath35 , @xmath36 and @xmath37 be the pointwise evaluation defined by @xmath38 . by hypothesis on each @xmath25 ,
@xmath37 is a linear form which implies that @xmath39 is also a linear form on @xmath40 .
now , for any @xmath12 that can be written as @xmath41 for @xmath42 , there exists a unique @xmath43 minimizing the functional @xmath44 and satisfying @xmath45 .
this unique element is given by the projection theorem for hilbert space @xcite @xmath46 which is a closed non - empty subspace in @xmath40 .
therefore @xmath2 is isometric to @xmath47 and hence it is a hilbert space . in order to identify the kernel of @xmath2
, we apply the lagrange multiplier rule .
an optimal @xmath43 corresponds to a stationary point of the augmented functional on @xmath48 @xmath49 remark that the norm @xmath50 makes the injection @xmath51 continuous and as a consequence @xmath52 is defined by duality .
therefore the pairing @xmath53 is well defined in formula .
then , at a stationary point @xmath54 we have @xmath55 note that in the previous formula , we could have written the heavier notation @xmath56 to be more precise . taking the dual pairing of the last equation with @xmath29 we get @xmath57 which implies that the riesz isomorphism between @xmath48 and @xmath2 is given by the map @xmath58 .
moreover , we have @xmath59 : for @xmath60 we have , latexmath:[\[\begin{gathered } which is true for any decomposition of @xmath12 so that @xmath62 since @xmath63 , @xmath2 is a rkhs and its kernel is @xmath64 .
we now define the isometric injection of @xmath2 in @xmath40 .
we denote by @xmath65 the map defined by @xmath66 .
the non - linear version of this multi - scale approach to diffeomorphic matching problems is the minimization of @xmath67 defined on @xmath68 .
recall that @xmath69 is the flow generated by @xmath70 .
the direct consequence of lemma [ reductionlemma ] is the following proposition .
[ equivalencematching ] the minimization of @xmath71 reduces to the minimization of @xmath72 obviously , the minimization of @xmath73 is the minimization of @xmath71 restricted to @xmath74 .
remark first that for any @xmath75,\bigoplus_{i=1}^k h_i)$ ] then @xmath76,\bigoplus_{i=1}^k h_i)$ ] . denoting @xmath77 we have @xmath78 with equality if and only if @xmath79 .
therefore , if @xmath75,\bigoplus_{i=1}^k h_i)$ ] is a minimizer of @xmath71 @xmath80 which implies @xmath79 and the result . to a minimizing path @xmath81,h)$ ] corresponds a minimizing path in @xmath68 via the map @xmath82 . in other words
, the optimal path can be decomposed on the different scales using each kernel @xmath34 present in the reproducing kernel of @xmath2 , @xmath83 . until now
, the scales have been introduced only on the lie algebra of the diffeomorphism group and an important question is how to decompose the flow of diffeomorphisms according to these scales .
an answer in the case of two scales is given in @xcite , where the flow is decomposed with the help of a semi - direct product of groups and the whole transformation is encoded in a large - scale deformation and a small - scale one .
the underlying idea is to represent the diffeomorphism flow @xmath84 by a pair @xmath85 where @xmath86 is given by the flow of the vector field @xmath87 and @xmath88 .
remark in particular that @xmath89 is not the flow of @xmath90 .
more precisely , we have @xmath91 interestingly , as shown in @xcite , this decomposition of the diffeomorphism flow corresponds to a semi - direct product of group .
this framework can be generalized to a finite number of scales as follows .
given @xmath92 scales , we want to represent @xmath93 by an @xmath92-tuple @xmath94 , with @xmath95 corresponding to the finest scale and @xmath96 to the coarsest scale .
the geometric construction underlying the decomposition into multiple scales is the semidirect product , introduced in the following lemma .
[ lem_sdp1 ] let @xmath97 be a chain of lie groups .
the @xmath92-fold semidirect product is given by the multiplication @xmath98 with @xmath99 denoting conjugation .
then given a right - invariant velocity field @xmath100 the curve @xmath101 is reconstructed via @xmath102 if @xmath103 and @xmath104 which is the semidirect product equivalent of @xmath105 .
we shall denote this semidirect product by @xmath106 to emphasize that each subproduct @xmath107 is a normal subgroup of the whole product .
verifying the axioms for the group multiplication is a straigt forward , if slightly longer calculation .
the inverse is given by @xmath108 the right hand side of equation can be obtained by differentiation the group multiplication at the identity , _
i.e. _ computing @xmath109 with @xmath110 fixed , @xmath111 and @xmath112 . in our case
the group @xmath113 is the diffeomorphism group @xmath114 generated by vector fields in the space @xmath115 corresponding to the kernel @xmath116 .
the subgroup condition @xmath117 is satisfied , if we impose the corresponding condition @xmath118 on the spaces of vector fields , which we will assume from now on .
starting from an @xmath92-tuple @xmath119 of vector fields , we can reconstruct the diffeomorphisms at each scale via @xmath120 as in lemma [ lem_sdp1 ] .
we can also sum over all scales to form @xmath121 and compute the flow @xmath84 of @xmath122 .
then a simple calculation shows that @xmath123 this construction can be summarized by the following commuting diagram @xmath124^<>(0.5){v(t ) = \sum v_k(t)}\ar[d]_{\text{via \eqref{rec_sdp1 } } } & & & v(t ) \ar[d]^{\p_t \ph(t ) = v(t ) \o \ph(t ) } \\ ( \ps_1(t),\ldots,\ps_n(t))\ar[rrr]_<>(0.5){\ph(t)= \ps_1(t ) \o \dots
\o \ps_n(t ) } & & & \ph(t ) } } \ ] ] we can now formulate several equivalent versions of lddmm matching with multiple scales . the most straight - forward way is to do matching with a kernel , which is a sum of kernels of different scales .
this is the approach considered in @xcite .
[ lddmm_sumofker ] registering @xmath125 to @xmath126 is done by finding the minimizer @xmath122 of @xmath127 where @xmath128 is a sum of @xmath92 kernels .
the corresponding simultaneous multiscale registration problem , where one assigns to each scale a separate vector field is a special case of the kernel bundle method proposed in @xcite and @xcite .
[ lddmm_simmultscale ] registering @xmath125 to @xmath126 is done by finding the minimizing @xmath92-tuple @xmath129 of @xmath130 where @xmath84 is the flow of the vector field @xmath131 .
the geometric version of the multiscale registration not only uses separate vector fields for each scale , but also decomposes the diffeomorphisms according to scales and can be defined as follows .
[ lddmm_sdp1 ] registering @xmath125 to @xmath126 is done by finding the minimizing @xmath92-tuple @xmath129 of @xmath130 where @xmath132 and @xmath133 is defined via .
problem [ lddmm_sdp1 ] can be obtained from the abstract framework in @xcite by considering the action @xmath134 of the semidirect product on the space of images .
the matching problems [ lddmm_sumofker ] , [ lddmm_simmultscale ] and [ lddmm_sdp1 ] are all equivalent .
the equivalence of problems [ lddmm_sumofker ] and [ lddmm_simmultscale ] follows from proposition [ equivalencematching ] while the equivalence of problems [ lddmm_simmultscale ] and [ lddmm_sdp1 ] follows from the diagram . for the case @xmath135 the proof can be found in more details in @xcite .
this construction can also be generalized to a continuum of scales as introduced in @xcite and @xcite .
we present in the next section a more general framework to deal with such continuous multi - scale decompositions .
the action of the semidirect product from lemma [ lem_sdp1 ] proceeds by deforming the image with the coarsest scale diffeomorphism first and with the finest scale diffeomorphism last . however , it is also possible to reverse this order and to act with the finest scale diffeomorphisms first .
we will see that this approach also correpsonds to a semidirect product and is equivalent to the other ordering of scales via a change of coordinate .
the reason to expand on this here is that it is better suited to be generalized to a continuum of scales . in this section
we will assume that the group @xmath136 contains the deformations of the coarsest scale and @xmath137 those of the finest scale .
the corresponding semidirect product is described in the following lemma .
[ lem_sdp2 ] let @xmath138 be a chain of lie groups .
the @xmath92-fold semidirect product is given by the multiplication @xmath139 with @xmath99 denoting conjugation .
then given a right - invariant velocity field @xmath100 the curve @xmath101 is reconstructed via @xmath140 which is the semidirect product equivalent of @xmath105 .
we shall denote this product by @xmath141 to emphasize that each subproduct @xmath142 from the left is a normal subgroup of the whole product .
this lemma can be proven in the same way as lemma [ lem_sdp1 ] .
these semidirect products defined in lemmas [ lem_sdp1 ] and [ lem_sdp2 ] are equivalent as shown by the following lemma the map @xmath143 is a group homomorphism between the two semidirect products and its derivative at the identity is given by @xmath144 direct computation .
the map @xmath145 can be seen as one side of the following commuting triangle @xmath146^\ph \ar[dr]_{t_1 } & & g_n \ltimes \dots\ltimes g_1 \ar[dl]^{t_2 } \\ & g_1\times\dots\times g_n & } \ ] ] the maps @xmath147 are group homomorphisms from the corresponding semidirect products into the direct product @xmath148 .
they can be regarded as trivializations of the semidirect product in the special case that the factors form a chain of subgroups .
we will now assume that we are given @xmath92 kernels @xmath149 such that the inclusions @xmath150 , _
i.e. _ @xmath151 represents the coarsest scale and @xmath152 the finest one .
note that the inclusions are reversed as compared to section [ sec_sdp ] .
the registration problem is now defined as follows .
[ lddmm_sdp2 ] registering @xmath125 to @xmath126 is done by finding the minimizing @xmath92-tuple @xmath129 of @xmath130 where @xmath132 and @xmath153 is defined via @xmath154 to see that problems [ lddmm_sdp1 ] and [ lddmm_sdp2 ] are equivalent note that the following diagram commutes @xmath146^\ph \ar[dr]_{\ph=\ps_1\dots\ps_n } & & g_n \ltimes \dots\ltimes g_1 \ar[dl]^{\ph=\ps_n\dots\ps_1 } \\ &
g_1 & } \ ] ] and that @xmath155 merely reverses the order of the vector fields @xmath129 .
we will see in section [ scale_decomposition ] that this version of the semidirect product is better suited to be generalized to a continuum of scales .
in this section , we define the multi - scale approach for a continuum of scales .
first , we introduce the necessary analytical framework and useful results .
[ def_admbundle ] let @xmath0 be a domain in @xmath156 .
an _ admissible bundle _
@xmath157 is a one - parameter family of reproducing kernel hilbert spaces of @xmath3 vector fields @xmath158 indexed by @xmath159 such that for any @xmath160 , there exists a positive constant @xmath161 s.t . for every @xmath162 @xmath163
the map @xmath164 is assumed to be borel measurable .
in addition to this set of linear spaces , @xmath165 is a borel measure on @xmath166 such that @xmath167 where @xmath168 is also assumed to be borel measurable . [ remoncontinuumsetting ]
* the hypothesis of the definition may not be optimal to obtain the needed property , but this context is already large enough for applications . * remark that no inclusion is a priori required between the linear spaces @xmath158 .
however the typical example is given by the usual scale - space _ i.e. _ @xmath169 defined by its gaussian kernel @xmath170 .
in this case , there exists an obvious towering @xmath171 for @xmath172 .
* we have @xmath173 .
this comes from cauchy - schwarzs inequality to get @xmath174 and the fact that @xmath175 .
mimicking section [ finitenumberscales ] , we consider the set of vector - valued functions defined on @xmath176 , namely denoting @xmath177 the lebesgue measure on @xmath0 , @xmath178 it is rather standard to prove that @xmath47 is a hilbert space for the norm introduced in .
remark that @xmath47 contains all the functions @xmath179 for any @xmath35 . directly from the assumptions on the space @xmath47 , we can define the set of functions @xmath180 then , the generalization of lemma [ reductionlemma ] @xcite reads in our situation : [ reductionlemmageneralization ] the space @xmath2 can be endowed with the following norm : for any @xmath181 , @xmath182 for @xmath183 satisfying the constraint @xmath184 .
this norm turns @xmath2 into a rkhs whose kernel is @xmath185 . in our case ,
the hypothesis on the bundle @xmath157 implies that @xmath2 is an admissible rkhs . as mentioned above ,
the map @xmath186 is well defined and so is @xmath187 .
using cauchy - schwarzs inequality , we have @xmath188 hence , the evaluation map @xmath189 is continuous on @xmath47
. therefore the map @xmath82 is continuous for the product topology on @xmath190 and its kernel is a closed subspace denoted by @xmath191 . applying the projection theorem on @xmath192 , for any @xmath193 there exists a unique @xmath194 such that @xmath195 .
hence , equation defines a norm on @xmath2 for which @xmath196 is an isometry .
in particular @xmath2 is a hilbert space .
remark that inequality means that @xmath2 is a rkhs .
we can now apply the lagrange multiplier rule in a hilbert space on the following functional @xmath197 where @xmath198 and @xmath199 . for a stationary point
, we obtain @xmath200 , @xmath201 , @xmath165 a.e .
the riesz isomorphism is then given by @xmath202 and the kernel function is given by @xmath203 importantly , the hypothesis on the bundle @xmath157 implies that @xmath2 is an admissible rkhs since we can apply a theorem of differentiation under the integral sign . by application of cauchy - schwarzs inequality to the integral defining @xmath204 for any @xmath199 , we have that @xmath205 in this section we will generalize the ideas of section [ sec_sdp ] from a finite sum of kernels to a continuum of scales
. we will make some more assumptions to the general setting introduced in section [ cont_math ] .
we assume that the measure @xmath206 is the lebesgue measure on the finite interval @xmath207 $ ] and that the family @xmath158 of rkhs is ordered by inclusion @xmath208 this assumption might be relaxed a little bit .
the main restriction is that we do nt want the measure @xmath209 to contain any singular parts .
as long as @xmath206 is absolutely continuous with respect to the lebesgue measure , _
i.e. _ it can be represented via a density @xmath210 , the same construction can be carried out .
the ordering of the inclusions corresponds to that in section [ sec_reverse ] .
as in the discrete setting we can formulate the two image matching problems .
the first is a direct generalization of problem [ lddmm_sumofker ] to a continuum of scales .
[ lddmm_intofker ] registering @xmath125 to @xmath126 is done by finding the minimizer @xmath122 of @xmath127 where @xmath211 is the integral over the scales @xmath212 .
the other problem associates to each scale a separate vector field .
it was proposed in @xcite , where it was called registration with a kernel bundle .
the term kernel bundle refers to the one - parameter family @xmath213}$ ] of rkhs .
[ lddmm_intsimmultscale ] registering @xmath125 to @xmath126 is done by finding the one - parameter family @xmath214 of vector fields , which minimizes @xmath215 where @xmath84 is the flow of the vector field @xmath216 .
these two problems are equivalent , as will be shown in theorem [ cont_equiv ] . as a next step
we want to obtain a geometric reformulation of the registration problem similar to [ lddmm_sdp1 ] or [ lddmm_sdp2 ] .
the goal of this reformulation is to decompose the minimizing flow of diffeomorphisms @xmath84 , such that the effect of each scale becomes visible . in order to do this decomposition
we define @xmath217 the following theorem allows us to interchange time and scale in the flow @xmath218 .
[ switch_st ] for each fixed @xmath219 , the one - parameter family @xmath220 is the flow of @xmath221 to prove this theorem we will use the following lemma .
[ switch_st_lemma ] let @xmath222 and @xmath223 be two - parameter families of vector fields which are @xmath224 in the @xmath225-variables and @xmath226 in @xmath227 . if they satisfy @xmath228(x)\ ] ] and if @xmath229 for all @xmath160 , then the flow of @xmath230 for fixed @xmath160 coincides with the flow of @xmath231 for fixed @xmath219
. denote by @xmath232 the flow of @xmath230 in @xmath219 .
then @xmath233\o a_s(t ) + du(s , t , a_s(t)).\p_sa_s(t ) \\ & = \p_t \left(v(s , t)\o a_s(t ) \right ) + du(s , t , a_s(t)).\left(\p_sa_s(t ) - v(s , t)\o a_s(t ) \right)\end{aligned}\ ] ] this implies that @xmath234 is the solution of the ode @xmath235 since for @xmath236 we have @xmath237 , it follows that @xmath238 is the unique solution of .
this means that @xmath239 _
i.e. _ the flows of @xmath230 in @xmath219 and of @xmath231 in @xmath160 coincide .
we apply lemma [ switch_st_lemma ] to the vector fields @xmath240 we can differentiate @xmath241 using the following rule @xmath242\enspace.\ ] ] using this we can verify the compatibility condition @xmath243 - v_s(t ) \\ & = \left[\int_0^s v_s(t ) { \,\mathrm{d}}r , \operatorname{ad}_{\ps_s(t ) } \int_0^t \operatorname{ad}_{\ps_s(r)\i }
v_s(r ) { \,\mathrm{d}}\right ] \enspace .\end{aligned}\ ] ] the condition @xmath229 is trivially satisfied .
this concludes the proof .
the main conclusion from theorem [ switch_st ] can be summarized in the corollary .
[ scale_flow ] let @xmath84 be the flow of the vector field @xmath244 and define @xmath245 via the left - invariant velocity , _
@xmath246 then we have @xmath247 _ i.e. _ @xmath245 is the flow in scale of @xmath84
. apply theorem [ switch_st ] with @xmath248 .
corollary [ scale_flow ] gives us a way to decompose the matching diffeomorphism @xmath249 into separate scales .
as we follow the flow @xmath245 , we add more and more scales , starting from the identity , when no scales are taken into account and finishing with @xmath249 , which includes all scales . in this sense @xmath250
contains the scale information for the scales in the interval @xmath251 $ ] .
it is of interest to understand the relation between a continuum of scales and the case , where we have only a finite number .
we will see , that it is possible to see the discrete case as an approximation of the conitnuous case .
let us start with a family @xmath212 of kernels with @xmath252 $ ] , where the scales are ordered from the coarsest to the finest , _
i.e. _ @xmath253 for @xmath254 as before .
divide the interval @xmath207 $ ] into @xmath92 parts @xmath255 and denote the intervals @xmath256 $ ] .
define the space @xmath257 where @xmath258 was defined in definition [ def_admbundle ] to be a one - parameter family of vector fields @xmath259 such that @xmath260 . to each interval
@xmath261 corresponds a kernel @xmath262 .
the discrete sampling map @xmath263 discretizes @xmath259 into @xmath92 scales .
formally we can introduce a lie bracket on the space @xmath264 by defining @xmath265_s = \left[u_s , \int_0^s v_r { \,\mathrm{d}}r\right ] + \left[\int_0^s u_r { \,\mathrm{d}}r , v_s\right]\enspace.\ ] ] using this bracket the sampling map @xmath266 is a lie algebra homomorphism as shown in the next theorem .
the sampling map @xmath266 is a lie algebra homomorphism from the lie algebra @xmath264 with the bracket defined in into the @xmath92-fold semidirect product with the bracket @xmath267 = ( [ u_1,v_1 ] , \ldots , [ u_k,\sum_{i=1}^{k-1 } v_i ] + [ \sum_{i=1}^{k-1 } u_i , v_k ] + [ u_k , v_k ] , \ldots ) \ ] ] using the definitions we first compute @xmath268_k & = \left[\int_{t_{k-1}}^{t_k } u_s { \,\mathrm{d}}s , \int_0^{t_{k-1 } } v_s { \,\mathrm{d}}s\right ] + \left[\int_{0}^{t_{k-1 } } u_s { \,\mathrm{d}}s , \int_{t_{k-1}}^{t_{k } } v_s { \,\mathrm{d}}s\right ] + \\ & \phantom{=\quad } + \left[\int_{t_{k-1}}^{t_k } u_s { \,\mathrm{d}}s , \int_{t_{k-1}}^{t_{k } } v_s { \,\mathrm{d}}s\right ] \\ & = \left[\int_{0}^{t_k } u_s { \,\mathrm{d}}s , \int_0^{t_{k } } v_s { \,\mathrm{d}}s\right ] - \left[\int_{0}^{t_{k-1 } } u_s { \,\mathrm{d}}s , \int_0^{t_{k-1 } } v_s { \,\mathrm{d}}s\right ] \enspace , \end{aligned}\ ] ] and then write the other side @xmath269)_k & = \int_{t_{k-1}}^{t_k } \left[u_s , \int_0^s v_r { \,\mathrm{d}}r\right ] + \left[\int_0^s u_r { \,\mathrm{d}}r , v_s\right ] { \,\mathrm{d}}s\end{aligned}\ ] ] below we interchange the order of integration in the first summand and merely switch @xmath160 and @xmath270 in the second summand to obtain @xmath271 + \left[\int_0^s u_r { \,\mathrm{d}}r , v_s\right ] { \,\mathrm{d}}s = \\ & = \int_0^{t_k } \int_0^s [ u_s , v_r ] { \,\mathrm{d}}r { \,\mathrm{d}}s + \int_0^{t_k } \int_0^r [ u_s , v_r ] { \,\mathrm{d}}s { \,\mathrm{d}}r \\ & = \int_0^{t_k } \int_r^{t_k } [ u_s , v_r ] { \,\mathrm{d}}s { \,\mathrm{d}}r + \int_0^{t_k } \int_0^r [ u_s , v_r ] { \,\mathrm{d}}s { \,\mathrm{d}}r\\ & = \int_0^{t_k } \int_0^{t_k } [ u_s , v_r ] { \,\mathrm{d}}s { \,\mathrm{d}}r \\ & = \left [ \int_0^{t_k } u_s { \,\mathrm{d}}s , \int_0^{t_k } v_s { \,\mathrm{d}}s \right ] \enspace.\end{aligned}\ ] ] decomposing the integral into @xmath272)_k = \int_0^{t_k } \dots { \,\mathrm{d}}s - \int_0^{t_{k-1 } } \dots { \,\mathrm{d}}s\ ] ] finishes the proof .
now we can show that all matching problems that we defined in the continuous case are equivalent .
[ cont_equiv ] the matching problems [ lddmm_intofker ] and [ lddmm_intsimmultscale ] are equivalent and using the sampling map @xmath266 they are also equivalent to the discrete problem [ lddmm_sdp2 ] .
the first equivalence follows from theorem [ reductionlemmageneralization ] . for the second equivalence note
that problem [ lddmm_sdp2 ] is equivalent to [ lddmm_sumofker ] and use lemma [ reductionlemma ] .
the diffeomorphisms @xmath273 at each scale , that were defined in [ lddmm_sdp2 ] are also contained in the continuous setting .
note that @xmath274 was the flow of the vector field @xmath275 and using the map
@xmath266 we have @xmath276 hence we obtain the identity @xmath277 , where @xmath278 was defined in .
in particular we retrieve @xmath279 the scale decomposition of the discrete case as a continuous scale descomposition evaluated at specific points .
in this paper , we have extended the mixture of kernels presented in @xcite to the continuous setting and we have given a variational interpretation of the matching problem .
in particular , we have shown that the approaches presented in @xcite and the mixture of kernels of @xcite are equivalent .
motivated by the mathematical development of the multi - scale approach to group of diffeomorphisms , we have extended the semi - direct product result of @xcite to a finite number of scales and also to a continuum of scales .
developing statistics in a multi - scale framework for the initial momentum is encouraged by previous results on the statistical use of each scales in @xcite . to this aim
, it would be very interesting question to obtain results about the sparsity of the scales used in the mixture of kernels in order to improve the statistical power on the initial momentum . in analogy with @xcite but from the image registration point of view
, @xcite tends to favor a non - sparse description of the scales , which tends towards the continuum of scales . in this direction , a more theoretical approach to learning the parameters involved in the mixture of kernels still needs to be developed . | we develop a multi - scale theory for group of diffeomorphisms based on previous works @xcite .
the purpose of the paper is to develop in details a variational approach for multi - scale on diffeomorphisms and to generalize to several scales the semi - direct product of group representation .
we also show that the approaches presented in @xcite and the mixture of kernels of [ rvw+11 ] are equivalent . |
in computer graphics programming the standard framework for modeling points in space is via a projective representation .
so , for handling problems in three - dimensional geometry , points in euclidean space @xmath0 are represented projectively as rays or vectors in a four - dimensional space , @xmath1 the additional vector @xmath2 is orthogonal to @xmath0 , @xmath3 , and is normalised to 1 , @xmath4 . from the definition of @xmath5 it is apparent that @xmath2 is the projective representation of the origin in euclidean space .
the projective representation is _ homogeneous _
, so both @xmath5 and @xmath6 represent the same point .
projective space is also not a linear space , as the zero vector is excluded . given a vector @xmath7 in projective space , the euclidean point @xmath8 is then recovered from @xmath9 the components of @xmath7 define a set of homogeneous coordinates for the position @xmath8 .
the advantage of the projective framework is that the group of euclidean transformations ( translations , reflections and rotations ) is represented by a set of linear transformations of projective vectors .
for example , the euclidean translation @xmath10 is described by the matrix transformation @xmath11 this linearisation of a translation ensures that compounding a sequence of translations and rotations is a straightforward exercise in projective geometry .
all one requires for applications is a fast engine for multiplying together @xmath12 matrices .
the main operation in projective geometry is the _ exterior product _
, originally introduced by grassmann in the nineteenth century @xcite .
this product is denoted with the wedge symbol @xmath13 .
the outer product of vectors is associative and totally antisymmetric .
so , for example , the outer product of two vectors @xmath7 and @xmath14 is the object @xmath15 , which is a rank-2 antisymmetric tensor or _ bivector_. the components of @xmath15 are @xmath16 the exterior product defines the _ join _ operation in projective geometry , so the outer product of two points defines the line between them , and the outer product of three points defines a plane . in this scheme a line in three dimensions
is then described by the 6 components of a bivector .
these are the plcker coordinates of a line .
the associativity and antisymmetry of the outer product ensure that @xmath17 which imposes a single quadratic condition on the coordinates of a line .
this is the plcker condition .
the ability to handle straight lines and planes in a systematic manner is essential to practically all graphics applications , which explains the popularity of the projective framework .
but there is one crucial concept which is missing .
this is the euclidean _ distance _ between points .
distance is a fundamental concept in the euclidean world which we inhabit and are usually interested in modeling . but distance can not be handled elegantly in the projective framework , as projective geometry is non - metrical .
any form of distance measure must be introduced via some additional structure .
one way to proceed is to return to the euclidean points and calculate the distance between these directly .
mathematically this operation is distinct from all others performed in projective geometry , as it does not involve the exterior product ( or duality ) .
alternatively , one can follow the route of classical planar projective geometry and define the additional metric structure through the introduction of the _ absolute conic _ @xcite . but
this structure requires that all coordinates are complexified , which is hardly suitable for real graphics applications .
in addition , the generalisation of the absolute conic to three - dimensional geometry is awkward . there is little new in these observations .
grassmann himself was dissatisfied with an algebra based on the exterior product alone , and sought an algebra of points where distances are encoded in a natural manner .
the solution is provided by the _
conformal model _ of euclidean geometry , originally introduced by mbius in his study of the geometry of spheres .
the essential new feature of this space is that it has mixed signature , so the inner product is not positive definite .
in the nineteenth century , when these developments were initiated , mixed signature spaces were a highly original and somewhat abstract concept . today
, however , physicists and mathematicians routinely study such spaces in the guise of special relativity , and there are no formal difficulties when computing with vectors in these spaces . as a route to understanding the conformal representation of points in euclidean geometry
we start with a description of the _
stereographic projection_. this map provides a means of representing points as null vectors in a space of two dimensions higher than the euclidean base space .
this is the conformal representation .
the inner product of points in this space recovers the euclidean distance , providing precisely the framework we desire .
the outer product extends the range of geometric primitives from projective geometry to include circles and spheres , which has many applications .
the conformal model of euclidean geometry makes heavy use of both the interior and exterior products . as such
, it is best developed in the language of _ geometric algebra _ a universal language for geometry based on the mathematics of _ clifford algebra
_ @xcite .
this is described in section [ sga ] .
the power of the geometric algebra development becomes apparent when we discuss the group of conformal transformations , which include euclidean transformations as a subgroup .
as in the projective case , all euclidean transformations are linear transformations in the conformal framework . furthermore , these transformations are all _ orthogonal _ , and can be built up from primitive reflections .
the join operation in conformal space generalises the join of projective geometry .
three points now define a line , which is the circle connecting the points .
if this circle passes through the point at infinity it is a straight line .
similarly , four points define a sphere , which reduces to a plane when its radius is infinite .
these new geometric primitives provide a range of intersection and reflection operations which dramatically extend the available constructions in projective geometry .
for example , reflecting a line in a sphere is encoded in a simple expression involving a pair of elements in geometric algebra .
working in this manner one can write computer code for complicated geometrical operations which is robust , elegant and highly compact .
this has many potential applications for the graphics industry .
the stereographic projection provides a straightforward route to the principle construction of the conformal model the representation of a point as a null vector in conformal space .
the stereographic projection maps points in the euclidean space @xmath18 to points on the unit sphere @xmath19 , as illustrated in figure [ fstereog ] .
suppose that the initial point is given by @xmath20 , and we write @xmath21 where @xmath22 is the magnitude of the vector @xmath0 , @xmath23 .
the corresponding point on the sphere is @xmath24 where @xmath25 is the unit vector perpendicular to the plane defining the south pole of the sphere @xmath19 .
the angle @xmath26 , @xmath27 is related to the distance @xmath22 by @xmath28 which inverts to give @xmath29 the stereographic projection @xmath30 maps @xmath18 into @xmath31 , where @xmath32 is the south pole of @xmath19 .
we complete the space @xmath19 by letting the south pole represent the point at infinity .
we therefore expect that , under euclidean transformations of @xmath18 , the point at infinity should remain invariant .
is mapped to the unit sphere @xmath19 . given a point in @xmath18 we form the line through this point and the south pole of the sphere .
the point where this line intersects the sphere defines the image of the projection.,width=340 ] we now have a representation of points in @xmath18 with unit vectors in the space @xmath33 . but the constraint that the vector has unit magnitude means that this representation is not homogeneous .
a homogeneous representation of geometric objects is critical to the power of projective geometry , as it enables us to write the equation of the line through @xmath7 and @xmath14 as @xmath34 clearly , if @xmath5 satisfies this equation , then so to does @xmath6 . to achieve a homogeneous representation we introduce a further vector , @xmath35 , which has _ negative signature _
, @xmath36 we also assume that @xmath35 is orthogonal to @xmath0 and @xmath25 .
we can now replace the unit vector @xmath37 with the vector @xmath5 , where @xmath38 the vector @xmath5 satisfies @xmath39 so is _
null_. this equation is homogeneous , so we can now move to a homogeneous encoding of points and let @xmath5 and @xmath6 represent the _ same _ point in @xmath18 . multiplying the vector in equation ( [ ex1 ] ) by @xmath40
we establish the conformal representation @xmath41 the vectors @xmath25 and @xmath35 extend the euclidean space @xmath18 to a space with two extra dimensions and signature @xmath42 .
it is generally more convenient to work with a null basis for the extra dimensions , so we define @xmath43 these vectors satisfy @xmath44 the vector @xmath5 is now @xmath45 which defines our standard representation of points as vectors in conformal space . given a general , unnormalised null vector in conformal space , the standard form of equation ( [ eptmap ] ) is recovered by setting @xmath46 this map makes it clear that the null vector @xmath47 now represents the point at infinity .
in general we will not assume that our points are normalised , so the components of @xmath5 are homogeneous coordinates for the point @xmath0 .
the step of normalising the representation is only performed if the actual euclidean point is required .
given two null vectors @xmath5 and @xmath48 , in the form of equation ( [ eptmap ] ) , their inner product is @xmath49 this is the key result which justifies the conformal model approach to euclidean geometry .
the inner product in conformal space encodes the _ distance _ between points in euclidean space .
this is why points are represented with null vectors the distance between a point and itself is zero . since equation ( [ econfinner ] ) was appropriate for normalised points , the general expression relating euclidean distance to the conformal inner product is @xmath50 this is manifestly homogeneous in @xmath5 and @xmath48 .
this formula returns the dimensionless distance . to introduce dimensions
one requires a fundamental length scale @xmath51 , so that @xmath0 is the dimensionless representation of the position vector @xmath52 .
appropriate factors of @xmath51 can then be inserted when required . an orthogonal transformation in conformal space
will ensure that a null vector remains null .
such a transformation therefore maps points to points in euclidean space .
this defines the full conformal group of euclidean space , which is isomorphic to the group @xmath53 .
conformal transformations leave angles invariant , but can map straight lines into circles .
the euclidean group is the subgroup of the conformal group which leaves the euclidean distance invariant .
these transformations include translations and rotations , which are therefore linear , orthogonal transformations in conformal space .
the key to developing simple representations of conformal transformations is geometric algebra , which we now describe .
the language of geometric algebra can be thought of as clifford algebra with added geometric content .
the details are described in greater detail elsewhere @xcite , and here we just provide a brief introduction .
a geometric algebra is constructed on a vector space with a given inner product .
the _ geometric _ product of two vectors @xmath8 and @xmath54 is defined to be associative and distributive over addition , with the additional rule that the ( geometric ) square of any vector is a scalar , @xmath55 if we write @xmath56 we see that the symmetric part of the geometric product of any two vectors is also a scalar .
this defines the inner product , and we write @xmath57 the geometric product of two vectors can now be written @xmath58 where the exterior product is the antisymmetric combination @xmath59 under the geometric product , orthogonal vectors anticommute and parallel vectors commute .
the product therefore encodes the basic geometric relationships between vectors .
the totally antisymmetrised sum of geometric products of vectors defines the exterior product in the algebra .
once one knows how to multiply together vectors it is a straightforward exercise to construct the entire geometric algebra of a vector space .
general elements of this algebra are called _ multivectors _ , and they too can be multiplied via the geometric product . the two algebras which concern us in this paper are the algebras of conformal vectors for the euclidean plane and three - dimensional space @xcite . for the euclidean plane ,
let @xmath60 denote an orthonormal basis set , and write @xmath61 , @xmath62 .
the full basis set is then @xmath63 .
these generators satisfy @xmath64 where @xmath65
the algebra generated by these vectors consists of the set @xmath66 scalars are assigned grade zero , vectors grade one , and so on .
the highest grade term in the algebra is the pseudoscalar @xmath67 , @xmath68 this satisfies @xmath69 the steps in this calculation simply involve counting the number of sign changes as a vector is anticommuted past orthogonal vectors .
this is essentially how all products in geometric algebra are calculated , and it is easily incorporated into any programming language . in even dimensions , such as the case here , the pseudoscalar anticommutes with vectors , @xmath70 for the case of the conformal algebra of euclidean three - dimensional space @xmath71 , we can define a basis as @xmath72 , with @xmath61 , @xmath73 , and @xmath74 a basis set for three - dimensional space .
the algebra generated by these vectors has 32 terms , and is spanned by @xmath75 the dimensions of each subspace are given by the binomial coefficients .
each subspace has a simple geometric interpretation in conformal geometry .
the pseudoscalar for five - dimensional space is again denoted by @xmath67 , and this time is defined by @xmath76 in five - dimensional space the pseudoscalar commutes with all elements in the algebra .
the @xmath77 signature of the space implies that the pseudoscalar satisfies @xmath78 so , algebraically , @xmath67 has the properties of a unit imaginary , though in geometric algebra it plays a definite geometric role . in a general geometric algebra , multiplication by
the pseudoscalar performs the duality transformation familiar in projective geometry .
suppose that the vectors @xmath8 and @xmath79 represent two lines from a common origin in euclidean space , and we wish to reflect the vector @xmath8 in the hyperplane perpendicular to @xmath79 .
if we assume that @xmath79 is normalised to @xmath80 , the result of this reflection is @xmath81 this is the standard expression one would write down without access to the geometric product . but with geometric algebra at our disposal we can expand @xmath82 into @xmath83 the advantage of this representation of a reflection is that we can easily chain together reflections into a series of geometric products .
so two reflections , one in @xmath79 followed by one in @xmath84 , produce the transformation @xmath85 but two reflections generate a rotation , so a rotation in geometric algebra can be written in the simple form @xmath86 where @xmath87 the tilde denotes the operation of _ reversion _ , which reverses the order of vectors in any series of geometric products .
given a general multivector @xmath7 we can decompose it into terms of a unique grade by writing @xmath88 where @xmath89 denotes the grade-@xmath22 part of @xmath7 .
the effect of the reverse on @xmath7 is then @xmath90 the geometric product of an even number of positive norm unit vectors is called a rotor .
these satisfy @xmath91 and generate rotations .
a rotor can be written as @xmath92 where @xmath14 is a bivector .
the space of bivectors is therefore the space of generators of rotations .
these define a _ lie algebra _
, with the rotors themselves defining a _ lie group_. the action of this group on vectors is defined by equation ( [ erotn ] ) , so both @xmath93 and @xmath94 define the same rotation .
transformations of conformal vectors which leave the product of equation ( [ eucdist ] ) invariant correspond to the group of euclidean transformations in @xmath18 . to simplify the notation
, we let @xmath95 denote the geometric algebra of a vector space with signature @xmath96 . the euclidean spaces of interest to us therefore have the algebras @xmath97 and @xmath98 associated with them .
the corresponding conformal algebras are @xmath99 and @xmath100 respectively .
each euclidean algebra is a subalgebra of the associated conformal algebra . the operations which mainly interest us here are translations and rotations .
the fact that translations can be treated as orthogonal transformations is a novel feature of conformal geometry .
this is possible because the underlying orthogonal group is non - compact , and so contains null generators . to see how this allows us to describe a translation ,
consider the rotor @xmath101 where @xmath8 is a vector in the euclidean space , so that @xmath102 .
the bivector generator satisfies
@xmath103 so is null .
the taylor series for @xmath104 therefore terminates after two terms , leaving @xmath105 the rotor @xmath104 transforms the null vectors @xmath47 and @xmath106 into @xmath107 and @xmath108 as expected , the point at infinity remains at infinity , whereas the origin is transformed to the vector @xmath8 .
acting on a vector @xmath109 we similarly obtain @xmath110 combining these results we find that @xmath111 which performs the conformal version of the translation @xmath112 .
translations are handled as rotations in conformal space , and the rotor group provides a double - cover representation of a translation .
the identity @xmath113 ensures that the inverse transformation in conformal space corresponds to a translation in the opposite direction , as required .
similarly , as discussed above , a rotation in the origin in @xmath18 is performed by @xmath114 , where @xmath93 is a rotor in @xmath115 .
the conformal vector representing the transformed point is @xmath116 this holds because @xmath93 is an even element in @xmath115 , so must commute with both @xmath47 and @xmath106 .
rotations about the origin therefore take the same form in either space . but suppose instead that we wish to rotate about the point @xmath117 .
this can be achieved by translating @xmath8 to the origin , rotating , and then translating forward again . in terms of @xmath118
the result is @xmath119 the rotation is now controlled by the rotor @xmath120 , where @xmath121 the conformal model frees us up from treating the origin as a special point .
rotations about any point are handled with rotors in the same manner .
similar comments apply to reflections , though we will see shortly that the range of possible reflections is enhanced in the conformal model .
the euclidean group is a subgroup of the full conformal group , which consists of transformations which preserve angles alone .
this group is described in greater detail elsewhere @xcite .
the essential property of a euclidean transformation is that the point at infinity is invariant , so all euclidean transformations map @xmath47 to itself .
in the conformal model , points in euclidean space are represented homogeneously by null vectors in conformal space . as in projective geometry ,
a multivector @xmath122 encodes a geometric object in @xmath18 via the equations @xmath123 one result we can exploit is that @xmath124 is unchanged if @xmath125 , where @xmath93 is a rotor in @xmath126 .
so , if a geometric object is specified by @xmath127 via equation ( [ ecf1 ] ) , it follows that @xmath128 we can therefore transform the object @xmath127 with a general element of the full conformal group to obtain a new object . as well as translations and rotations , the conformal group includes dilations and inversions , which map straight lines into circles .
the range of geometric primitives is therefore extended from the projective case , which only deals with straight lines . the first case to consider is a pair of null vectors @xmath7 and @xmath14 .
their inner product describes the euclidean distance between points , and their outer product defines the bivector @xmath129 the bivector @xmath130 has magnitude @xmath131 which shows that @xmath130 is _ timelike _ , in the terminology of special relativity .
it follows that @xmath130 contains a pair of null vectors .
if we look for solutions to the equation @xmath132 the only solutions are the two null vectors contained in @xmath130 .
these are precisely @xmath7 and @xmath14 , so the bivector encodes the two points directly . in the conformal model ,
no information is lost in forming the exterior product of two null vectors .
frequently , bivectors are obtained as the result of intersection algorithms , such as the intersection of two circles in a plane .
the sign of the square of the resulting bivector , @xmath133 , defines the number of intersection points of the circles . if @xmath134 then @xmath14 defines two points , if @xmath135 then @xmath14 defines a single point , and if @xmath136 then @xmath14 contains no points .
given that bivectors now define pairs of points , as opposed to lines , the obvious question is how do we encode lines ?
suppose we construct the line through the points @xmath137 . a point on the line
is then given by @xmath138 the conformal version of this line is @xmath139 and any multiple of this encodes the same point on the line .
it is clear , then , that a conformal point @xmath5 is a linear combination of @xmath7 , @xmath14 and @xmath47 , subject to the constraint that @xmath124 .
this is summarised by @xmath140 so it is the _ trivector _
@xmath141 which represents a line in conformal geometry .
this illustrates a general feature of the conformal model geometric objects are represented by multivectors one grade higher than their projective counterpart .
the extra degree of freedom is absorbed by the constraint that @xmath124 .
now suppose that we form a general trivector @xmath127 from three null vectors , @xmath142 this must still encode a conformal line via the equation @xmath143 .
in fact , @xmath127 encodes the _ circle _ through the points defined by @xmath144 , @xmath145 and @xmath146 . to see why ,
consider the conformal model of a plane .
the trivector @xmath127 therefore maps to a _ dual _ vector @xmath84 , where @xmath147 and @xmath67 is the pseudoscalar defined in equation ( [ eps2 ] ) .
we see that @xmath148 so @xmath84 is a vector with positive square .
such a vector can always be written in the form @xmath149 where @xmath150 is the conformal vector for the point @xmath151 .
the dual version of the equation @xmath152 is @xmath153 which reduces to @xmath154 since @xmath150 satisfies @xmath155 , equation ( [ ecirc ] ) states that the euclidean distance between @xmath0 and @xmath156 is equal to the constant @xmath157 .
this clearly defines a circle in a plane .
furthermore , the radius of the circle is defined by @xmath158 where the minus sign in the final expression is due to the fact that the 4-vector @xmath159 has negative square .
this equation demonstrates how the conformal framework allows us to encode dimensional concepts such as radius while keeping multivectors like @xmath127 as homogeneous representations of geometric objects .
the equation for the radius of the circle tells us that the circle has infinite radius if @xmath160 this is the case of a straight line , and this equation can be interpreted as saying the line passes through the point at infinity .
so , given three points , the test that they lie on a line is @xmath161 this test has an important advantage over the equivalent test in projective geometry .
the degree to which the right - hand side differs from zero directly measures how far the points are from lying on a common line .
this can resolve a range of problems caused by the finite numerical precision of most computer algorithms .
numerical drift can only affect how near to being straight the line is . when it comes to plotting
, one can simply decide what tolerance is required and modify equation ( [ elntst ] ) to read @xmath162 where @xmath163 .
a similar idea can not be applied so straightforwardly in the projective framework , as there is no intrinsic measure of being ` nearly linearly dependent ' for projective vectors .
next , suppose that @xmath8 and @xmath54 define two vectors in @xmath18 , both as rays from the origin , and we wish to find the angle between these .
the conformal representation of the line can be built up from @xmath164 where @xmath165 and @xmath166 .
similarly the line in the @xmath54 direction is represented by @xmath167 .
we can therefore write @xmath168 so that @xmath169 if we let angle brackets @xmath170 denote the scalar part of the multivector @xmath171 we see that @xmath172 where @xmath26 is the angle between the lines .
this is true for lines through the origin , but the expression in terms of @xmath173 and @xmath174 is unchanged under the action of a general rotor , so applies to lines meeting at any point , and to circles as well as straight lines .
similar considerations apply to circles and planes in three dimensional space .
suppose that the points @xmath175 define a plane in @xmath18 , so that an arbitrary point in the plane is given by @xmath176 the conformal representation of @xmath0 is @xmath177 where @xmath178 _ etc .
_ , and @xmath179 varying @xmath180 and @xmath181 , together with the freedom to scale @xmath182 , now produces general null combinations of the vectors @xmath7 , @xmath14 , @xmath183 and @xmath47 . the equation for the plane
can then be written @xmath184 so , as one now expects , it is 4-vectors which define planes .
if instead we form the 4-vector @xmath185 then in general @xmath186 defines a sphere . to see why ,
consider the dual of @xmath30 , @xmath187 where @xmath67 is now given by equation ( [ eps3 ] ) .
again we find that @xmath188 , and we can write @xmath189 the equation @xmath190 is now equivalent to @xmath191 , which defines a sphere of radius @xmath157 , with centre @xmath150 . the radius of the sphere is defined by @xmath192 the sphere becomes a flat plane if the radius is infinite , so the test that four points lie on a common plane is @xmath193 as with the case of the test of collinearity , this is numerically well conditioned , as the deviation from zero is directly related to the curvature of the sphere through the four points .
the conformal model extends the range of geometric primitives beyond the simple lines and planes of projective geometry .
it also provides a range of new algorithms for handling intersections and reflections .
first , suppose we wish to find the intersection of two lines in a plane .
these are defined by the trivectors @xmath173 and @xmath174 .
the intersection , or meet , is defined via its dual by @xmath194 the star , @xmath195 , denotes the dual of a multivector , which in geometric algebra is formed by multiplication by the multivector representing the join . for the case of two distinct lines in a plane ,
their join is the plane itself , so the dual is formed by multiplication by the pseudoscalar @xmath67 of equation ( [ eps2 ] ) .
the intersection of two lines therefore results in the bivector @xmath196 this expression is a bivector as it is the contraction of a vector and a trivector . as discussed in section [ sprims ] , a bivector can encode zero , one or two points , depending on the sign of its square .
two circles can intersect in two points , for example , if they are sufficiently close to one another .
two straight lines will also intersect in two points , though one of these is at infinity . in three dimensions
a number of further possibilities arise .
the intersection of a line @xmath127 and a sphere or plane @xmath32 also results in a bivector , @xmath197 where @xmath67 is now the grade-5 pseudoscalar in the conformal algebra for three - dimensional euclidean space
. the bivector encodes the fact that a line can intersect a sphere in up to two places .
if @xmath32 is a flat plane and @xmath127 a straight line , then one of the intersection points will be at infinity .
more complex is the case of two planes or spheres . in this case
both geometric primitives are 4-vectors , @xmath198 and @xmath199 .
their intersection results in the trivector @xmath200 this trivector encodes a line .
this one single expression covers a wide range of situations , since either plane can be flat or spherical .
if both planes are flat their intersection is a straight line with @xmath201 . if one or both of the planes are spheres , their intersection results in a circle .
the sign of the square of the trivector @xmath127 immediately encodes whether or not the spheres intersect . if @xmath202 the spheres do not intersect , as there are no null solutions to @xmath203 when @xmath202 . as well as new intersection algorithms ,
the application of geometric algebra in conformal space provides a highly compact encoding of reflections . as a single example
, suppose that we wish to reflect the line @xmath127 in the plane @xmath32 . to start with ,
suppose that the point of intersection is the origin , with @xmath127 a straight line in the @xmath8 direction , and @xmath32 the plane defined by the points @xmath54 , @xmath156 and the origin .
in this case we have @xmath204 the result of reflecting @xmath8 in @xmath205 is a vector in the @xmath82 direction , where @xmath206 the conformal representation of the line through the origin in the @xmath82 direction is @xmath207 . in terms of their conformal representations , we have @xmath208 so far this is valid at the origin , but conformal invariance ensures that it holds for all intersection points .
this expression for @xmath209 finds the reflected line in space without even having to find the intersection point .
furthermore , the line can be curved , and the same expression reflects a circle in a plane
. we can also consider the case where the plane itself is curved into a sphere , in which case the transformation of equation ( [ eplnrf ] ) corresponds to an _ inversion _ in the sphere .
inversions are important operations in geometry , though perhaps of less interest in graphics applications . to find the reflected line for the spherical case
is also straightforward , as all one needs is the formula for the tangent plane to the sphere @xmath30 at the point of intersection .
this is @xmath210 where @xmath5 is the point where the line @xmath127 intersects the sphere .
the plane @xmath32 can then be used in equation ( [ eplnrf ] ) to find the reflected line .
conformal geometry provides an efficient framework for studying euclidean geometry because the inner product of conformal vectors is directly related to the euclidean distance between points .
the true power of the conformal framework only really emerges when the subject is formulated in the language of geometric algebra .
the geometric product unites the outer product , familiar from projective geometry , with the inner product of conformal vectors .
each graded subspace encodes a different geometric primitive , which provides for good data typing and aids writing robust code .
furthermore , the algebra includes multivectors which generate euclidean transformations .
so both geometric objects , and the transformations which act on them , are contained in a single , unified framework .
the remaining step in implementing this algebra in applications is to construct fast algorithms for multiplying multivectors .
much as projective geometry requires a fast engine for multiplying @xmath211 matrices , so conformal geometric algebra requires a fast engine for multiplying multivectors .
one way to achieve this is to encode each multivector via its matrix representation . for the important case of @xmath100 this representation consists of @xmath212 complex matrices .
but in general this approach is slower than algorithms designed to take advantage of the unique properties of geometric algebra .
such algorithms are under development by a number of groups around the world .
these will doubtless be of considerable interest to the graphics community .
cd would like to thank the organisers of the conference for all their work during the event , and their patience with this contribution .
cd is supported by an epsrc advanced fellowship .
lasenby and j. lasenby .
surface evolution and representation using geometric algebra . in r.
cippola , a. martin , editors , _ the mathematics of surfaces ix : proceedings of the ninth i m a conference on the mathematics of surfaces _ , pages 144168 .
london , 2000 . | projective geometry provides the preferred framework for most implementations of euclidean space in graphics applications .
translations and rotations are both linear transformations in projective geometry , which helps when it comes to programming complicated geometrical operations .
but there is a fundamental weakness in this approach the euclidean distance between points is not handled in a straightforward manner . here
we discuss a solution to this problem , based on conformal geometry .
the language of geometric algebra is best suited to exploiting this geometry , as it handles the interior and exterior products in a single , unified framework .
a number of applications are discussed , including a compact formula for reflecting a line off a general spherical surface . * conformal geometry , euclidean space + and geometric algebra * chris doran , anthony lasenby and joan lasenby cambridge university , uk keywords : geometric algebra , clifford algebra , conformal geometry , projective geometry , homogeneous coordinates , sphere geometry , stereographic projection |
in this section we motivate our study of the ( two - dimensional ) voter model and its dual coalescing walks through their connection with a number of percolation models . in section 2
, we report on numerical results for the dimension of a natural `` chordal interface '' of the voter model . in section 3
we give rigorous ( and a few numerical ) results on the large coalescing classes for coalescing walks ( where vertices @xmath2 and @xmath3 in a box are in the same class if their walks coalesce before hitting the boundary ) . in the appendix ,
more details about our numerical results are provided . among the most important breakthroughs in statistical physics and probability in the last two decades
is the work by schramm and coauthors @xcite and smirnov @xcite identifying ( or conjecturing ) members of the schramm - loewner evolution family of random curves as the scaling limits of various random walks and interfaces in two - dimensional spin systems .
in particular smirnov @xcite ( see also camia and newman s paper @xcite ) has shown that the scaling limit of critical site - percolation on the triangular lattice @xmath4 is sle@xmath5 . to give a rough description of one version of this statement ,
take a rhombic box @xmath6 ( containing @xmath7 vertices ) in the triangular lattice in two dimensions .
label the sides clockwise starting from the southwest corner as @xmath8 .
a percolation configuration on @xmath6 is an element @xmath9 of @xmath10 defined as follows .
fix the vertices in @xmath11 and @xmath12 to have value 0 ( or black or closed ) and those in @xmath13 and @xmath14 to be 1 ( or red or open ) . in the interior @xmath15 , set each vertex to be ( independently ) 0 or 1 , with probability @xmath16 each see figure [ fig : small_perc ] . with @xmath17 , with the exploration path shown in green . ]
there exists a unique simple path @xmath18 of length @xmath19 from the southwest corner following edges in the dual hexagonal lattice to the opposite corner that keeps black / closed vertices on the left and red / open vertices on the right .
@xmath20 is often referred to as the exploration path ; we will also call it the chordal interface . as @xmath21 the law of @xmath20 , after rescaling , converges weakly to a probability measure on continuous paths that is the law of _
chordal sle@xmath5 _ @xcite in a rhombic domain .
one can use this to prove ( see ( * ? ? ? * prop .
2 ) ) that @xmath22\approx l^{7/4}$ ] , where @xmath23 in general for @xmath24 we have that @xmath25 if and only if @xmath26 note that @xmath27 uniquely determines the path @xmath18 . one can therefore ask about the limiting behavior of @xmath20 when the configurations @xmath28 are generated by some other process ( i.e. , not i.i.d
. critical site percolation ) in the interior @xmath15 . in the case of the ising model ( where the states at two sites are not independent ) at the critical temperature ,
smirnov @xcite has identified that the limiting probability measure is instead chordal _
sle@xmath29_. we are interested in the limiting behavior of @xmath20 when the law of the configuration @xmath28 is the stationary distribution of the voter model ( or related models ) on @xmath15 . in this section
we define our primary model of interest , on @xmath6 as described above , with boundary states set as 0 on one pair of adjacent sides and 1 on the other pair , while the law of the interior states @xmath30 is the stationary measure for the voter model @xmath31 on @xmath15 , as follows .
each @xmath32 has its own independent poisson clock ( a poisson process @xmath33 ) of rate 1 . when the clock of a vertex @xmath34 rings we update the state @xmath35 of @xmath34 by choosing one of its six neighbors uniformly at random and adopting the state of the chosen neighbor .
note that the neighbor may be one of the vertices in the boundary @xmath36 whose state is fixed .
defined this way , @xmath37 is an irreducible markov process with finite state space @xmath38 , and therefore it has a unique invariant distribution .
we will write @xmath39 for a random configuration sampled from this invariant distribution .
the process admits a well - known graphical representation ( due to t.e .
harris @xcite ) which we now review .
for each @xmath40 , we draw a positive half line ( representing time ) in the third dimension , and on it we mark the times of poisson clock rings of that vertex .
each mark on a time line represents a state update event which also has an arrow from @xmath34 to the uniformly chosen neighbor whose state is adopted .
the lines of the boundary vertices have arrow marks to them , but not from them , as those states are fixed .
fix an initial configuration @xmath41 . to determine the state of a vertex @xmath42 at time @xmath43 we start at height / time @xmath43 on the time line corresponding to @xmath34 and
follow it down until we reach height / time 0 or we encounter an outgoing arrow ( whichever comes first ) at height @xmath44 .
if we meet an outgoing arrow we follow it to the time line of a neighboring vertex @xmath45 and proceed as before , following this time line down from height @xmath46 until reaching height / time 0 or an outgoing arrow .
we stop this procedure when we reach a boundary vertex or height 0 on some time line .
thus from any @xmath42 and @xmath47 the path followed corresponds to a continuous time nearest neighbor simple random walk on @xmath48 stopped upon reaching a boundary vertex or height 0 . in either case
the state @xmath49 at the terminal vertex is known and we set @xmath50 .
such a system of `` state genealogy walks '' from all the vertices at time @xmath43 following backward in time is a dual model and is distributed as a system of _ coalescing _ simple symmetric continuous time random walks on the triangular lattice see for example @xcite . since @xmath48 is finite ,
if @xmath43 is large enough all the walks starting then will with high probability hit the boundary before reaching height 0 . indeed ,
if we continue the time lines and poisson clocks below height 0 ( and do not terminate the walks at height 0 ) then almost surely from any height @xmath43 there will be a random height @xmath51 at which the walks started from all vertices @xmath42 at height @xmath43 will have reached boundary vertices .
what happens on the time lines below height @xmath52 does not affect @xmath53 since the states of the boundary vertices are fixed for all time .
this is equivalent to saying that the voter model itself reaches stationarity by a random finite time ( distributed as @xmath54 ) .
therefore to sample from @xmath39 it is enough to follow a system of coalescing continuous time simple random walks from each vertex @xmath34 of @xmath55 until they hit a boundary vertex @xmath56 , and set @xmath57 , i.e. @xmath58 if @xmath59 , and @xmath60 otherwise . one could instead sample from @xmath39 by setting @xmath61 for every @xmath62 ( so the state space would become @xmath63 ) and simulating the voter model dynamics until there is no vertex with state 2 .
figure [ fig:1sample ] shows a simulation of @xmath39 with @xmath64 , obtained by simulating coalescing random walks from each vertex in the interior , until each one has reached the boundary . with @xmath64 , simulated using c++
the white curve is the exploration path or chordal interface . ]
the duality discussed in the previous section tells us that @xmath65 ( we drop the superscript @xmath0 when there is no ambiguity ) is the probability that a simple random walk started at @xmath2 first hits the boundary at a 1-site . in other words ,
the one - dimensional distributions of our voter model on @xmath48 are equal to those of a model we would like to call _ harmonic percolation_. this is a model under which the states @xmath66 are independent of each other , and as we have already suggested , @xmath67 is equal to the probability that a simple random walk started from @xmath34 first hits the boundary @xmath68 at a 1-site ( i.e. , in @xmath69 ) .
harmonic percolation on an infinite strip of thickness @xmath0 coincides with an independent percolation model called gradient percolation @xcite . in the case of gradient percolation the probability @xmath70 of a site @xmath2 being open changes linearly from one boundary where it is 0 to the other boundary where it is 1 .
thus the function @xmath70 is harmonic inside the strip ( with specified boundary conditions ) .
the difference between the voter and harmonic percolation models arises from the fact that the walks in the former are coalescing , whereas in the latter they are independent . to be more explicit ,
coalescence in the voter model leads to non - zero correlations as in the following simple lemma .
[ lem : correlation ] for any @xmath71 , and any @xmath72 in the interior of @xmath48 , @xmath73 fix @xmath0 , @xmath74 and @xmath75 , and let @xmath76 and @xmath77 denote the elements of @xmath68 with fixed states 0 and 1 respectively .
let @xmath78 and @xmath79 be two independent random walks starting from @xmath74 and @xmath75 respectively .
let @xmath80 be the first time that @xmath78 and @xmath79 meet each other .
let @xmath81 for all times and define @xmath82 so that @xmath83 and @xmath84 are coalescing walks started from @xmath74 and @xmath75 respectively .
let @xmath85 and @xmath86 denote the respective hitting times of the boundary , and note that @xmath87 .
then @xmath88 this proves the result for @xmath89 .
a similar coupling argument can be made for any number @xmath90 of walkers starting from vertices @xmath91 ( choosing the lower indexed random walker to continue when any two meet ) , establishing the claim . for any @xmath92 , if @xmath93 and @xmath94 are distance at least @xmath95 from each other and the boundary @xmath96 then there exist @xmath97 and @xmath98 such that @xmath99 for @xmath100 and all @xmath0 , while @xmath101 where @xmath102 and @xmath103 are times when the difference random walk @xmath104 started at @xmath105 first hits the origin and the boundary of the box @xmath106 respectively , and the last equality follows from proposition 6.4.3 of @xcite .
then implies that the correlation @xmath107 between the votes at @xmath74 and @xmath75 goes to zero as per the following .
[ lem : corgoeszero ] let @xmath92 , and @xmath93 and @xmath94 be distance at least @xmath95 from each other and the boundary @xmath96 .
then @xmath108 as @xmath1
. one can consider i.i.d .
percolation , harmonic percolation , and the stationary voter model on @xmath48 as special cases of a general 2-parameter family of models as follows .
start a continuous - time walker from each site .
each walker initially wears a hat .
two walkers wearing hats coalesce when they meet , and instantly become a single walker wearing a hat .
walkers not wearing hats do not coalesce with any other walkers .
in addition a poisson clock is assigned to each walker .
when such a clock rings , the walker takes a random walk step , but before doing so removes her coalescence hat with probability @xmath109 .
if a walker wearing a hat steps into a site with another walker with a hat on , the walker that just made its step becomes part of the coalescence set of the walker that was already at the site . upon hitting a boundary site , with probability
@xmath110 a walker ( and its entire coalescence set ) is assigned the vote of the boundary vertex it hit , and otherwise ( i.e. , with probability @xmath111 ) its entire coalescence set attains an independently and uniformly chosen vote .
varying the boundary and coalescence noise parameters @xmath111 and @xmath109 between 0 and 1 allows us to interpolate between the four corner models : the voter model @xmath112 ; harmonic percolation @xmath113 ; i.i.d .
percolation @xmath114 ; and the case @xmath115 corresponds to a model we would like to call cow ( coalescing walk ) percolation .
recall that @xmath19 denotes the length of the interface . since this path is a nearest neighbor simple path
, there exist @xmath116 such that @xmath117 almost surely .
we conjecture that @xmath118\approx l^{d}\end{aligned}\ ] ] for some @xmath119 $ ] . in the case of critical i.i.d .
percolation , holds with @xmath120 which is also the hausdorff dimension of the limiting law ( i.e. , of sle@xmath5 , see @xcite ) . for gradient percolation on an infinite strip ,
the interface curve between the occupied cluster and empty cluster is a.s .
unique and has expected length approximately @xmath121 , where @xmath122 is the horizontal length of the piece of strip in which we measure boundary length ( * ? ? ?
* proposition 11 ) .
so , for any @xmath92 , for all sufficiently large @xmath0 , if we take a piece of strip which is @xmath0 long ( and @xmath0 thick ) , the expected length of the interface curve @xmath123 satisfies @xmath124 . for any @xmath125 , with probability going to 1 with @xmath0 ,
the curve stays in the central band ( around the central @xmath126 line where @xmath127 ) of width @xmath128 ( * ? ? ?
* theorem 6 ) .
thus , as @xmath1 , unless we appropriately zoom in around the central line , we expect to see the rescaled interface curve converge to a straight line in the center .
since the harmonic function inside a rhombic area with our boundary condition looks almost linear along the diagonal that connects the middle corner of the 1 valued boundary to the middle corner of the 0 valued boundary ( or indeed along any parallel line ) , we expect that the interface curve for harmonic percolation inside our rhombus should scale to a straight line as well .
writing @xmath129 for some function @xmath130 which makes the equality true we have that @xmath131 computing the average interface curve length @xmath132 from @xmath133 independent realizations of @xmath39 we obtain the following estimators for @xmath134 based on @xmath135 we say that an estimator @xmath136 ( more precisely a family of estimators @xmath137 ) is a _ consistent _ estimator of some quantity @xmath138 if @xmath139 it is easy to show that @xmath140 is a consistent estimator of @xmath134 if and only if @xmath141 as @xmath142 ( i.e. , if and only if @xmath143 ) , while @xmath144 is a consistent estimator for @xmath134 if and only if @xmath145 as @xmath1 . thus both estimators are consistent if @xmath146 is slowly varying at @xmath147 . if we are willing to assume that the random interface length @xmath19 in a box of size @xmath0 satisfies @xmath148 , where @xmath149=0 $ ] , then it is natural to consider the ordinary least squares estimator @xmath150 for the slope coefficient @xmath134 of the simple linear regression model @xmath151 where @xmath152 are interface lengths on boxes of side lengths @xmath153 , and the @xmath154 are random variables with mean 0 . note that @xmath130 is constant under this assumption .
the results from independent simulations of computing the average lengths of the interface curve and estimates @xmath155 ( with @xmath156 ) appear in table [ tab : d_estimates ] .
for ordinary percolation and harmonic percolation the values are known ( or expected ) to be @xmath157 and @xmath158 , so for these models @xmath144 ( with @xmath159 ) appears to do best . for the voter model
the value of @xmath144 is about 1.46 .
.estimates of @xmath134 with @xmath156 and with @xmath160 ( and @xmath161 ) , rounded to 4 decimal places . for harmonic percolation and percolation , @xmath144 is the closest estimator to the true values of 1.42857 and 1.75 respectively .
[ cols= " < , < , < , < , < , < , < , < , > " , ] | consider the voter model on a box of side length @xmath0 ( in the triangular lattice ) with boundary votes fixed forever as type 0 or type 1 on two different halves of the boundary .
motivated by analogous questions in percolation , we study several geometric objects at stationarity , as @xmath1 .
one is the interface between the ( large i.e. , boundary connected ) 0-cluster and 1-cluster .
another is the set of large `` coalescing classes '' determined by the coalescing walk process dual to the voter model . |
gene regulatory networks play a central role in cellular function by translating genotype into phenotype . by dynamically controlling gene expression ,
gene regulatory networks provide cells with a mechanism for responding to environmental challenges .
therefore , creating accurate mathematical models of gene regulation is a central goal of mathematical biology .
delay in protein production can significantly affect the dynamics of gene regulatory networks .
for example , delay can induce oscillations in systems with negative feedback @xcite , and has been implicated in the production of robust , tunable oscillations in synthetic gene circuits containing linked positive and negative feedback @xcite .
indeed , delayed negative feedback is thought to govern the dynamics of circadian oscillators @xcite , a hypothesis experimentally verified in mammalian cells @xcite . in genetic regulatory networks , noise and delay
interact in subtle and complex ways .
delay can affect the stochastic properties of gene expression and hence the phenotype of the cell @xcite .
it is well known that noise can induce switching in bistable genetic circuits @xcite ; the infusion of delay dramatically enhances the stability of such circuits @xcite and can induce an analog of stochastic resonance @xcite .
variability in the delay time ( distributed delay ) can accelerate signaling in transcriptional signaling cascades @xcite .
given the importance of delay in gene regulatory networks , it is necessary to develop methods to simulate and analyze such systems across spatial scales . in the absence of delay
, it is well known that chemical reaction networks are accurately modeled by ordinary differential equations ( odes ) in the thermodynamic limit , _
i.e. _ when molecule numbers are sufficiently large .
when molecule numbers are small , however , stochastic effects can dominate . in this case , the chemical master equation ( cme ) describes the evolution of the probability density function over all states of the system .
gillespie s stochastic simulation algorithm ( ssa ) @xcite samples trajectories from the probability distribution described by the cme .
while exact , the cme is difficult to analyze and the ssa can be computationally expensive . to address these issues ,
a hierarchy of coarse - grained approximations of the ssa has been developed @xcite ( see figure [ f : hierarchy ] ) .
spatially discrete approximations , such as @xmath0-leaping @xcite and @xmath1-leaping @xcite trade exactness for efficiency . at the next level are chemical langevin equations ( cles ) , which are stochastic differential equations of dimension equal to the number of species in the biochemical system .
cles offer two advantages .
first , unlike the ssa , the well - developed ideas from random dynamical systems and stochastic differential equations apply to cles .
second , it is straightforward to simulate large systems using cles
. finally , in the thermodynamic limit , one arrives at the end of the markovian hierarchy : the reaction rate equation ( rre ) .
( 1,0.75770602 ) ( 0,0 ) ( 0.63978037,0.66330034)(0,0)[lb ] ( 0.63353796,0.08899866)(0,0)[lb ] ( 0.5960835,0.23257399)(0,0)[lb ] ( 0.63353796,0.36366461)(0,0)[lb ] ( 0.38384251,0.65966796)(0,0)[lb ] ( 0.47656487,0.50123945)(0,0)[lb ] ( 0.50777693,0.33893741)(0,0)[lb ] ( 0.2331119,0.33893741)(0,0)[lb ] ( 0.30802054,0.19536251)(0,0)[lb ] ( 0.50777688,0.07051479)(0,0)[lb ] ( 0.63978037,0.66330034)(0,0)[lb ] ( 0.78959824,0.33869497)(0,0)[lb ] ( 0.78959824,0.07651419)(0,0)[lb ] ; see section [ sec : pf ] . ]
the markovian hierarchy above ( no delay ) is well - understood @xcite , but a complete analogue of the markovian theory does not yet exist for systems with delay .
the ssa has been generalized to a delay version - the dssa - to allow for both fixed @xcite and variable @xcite delay .
some analogues of @xmath2-leaping exist for systems with delay ; see _ e.g. _ @xmath3-leaping @xcite .
several methods have been used to formally derive a delay chemical langevin equation ( dcle ) from the delay chemical master equation ( dcme ) ; see section [ sec : discuss ] for details .
brett and galla @xcite use the path integral formalism of martin , siggia , rose , janssen , and de dominicis to derive a dcle approximation without relying on a master equation .
the brett and galla derivation produces the ` correct ' dcle approximation of the underlying delay birth - death ( dbd ) process in the sense that the first and second moments of the dcle match those of the dbd process .
however , their derivation has some limitations ( see section [ sec : discuss ] ) . in particular
, it gives no rigorous quantitative information about the distance between the dbd process and the dcle . in this paper
, we establish a rigorous link between dbd processes and dcles by proving that the distance between the dbd process and the correct approximating dcle process converges to zero as system size tends to infinity ( as measured by expectations of functionals of the processes ) .
in particular , this result applies to all moments .
it is natural to express distance in terms of expectations of functionals because the dbd process is spatially discrete while the correct dcle produces continuous trajectories ( see figure [ fig : intro ] ) .
further , we prove that both processes converge weakly to the thermodynamic limit .
finally , we quantitatively estimate the distance between the dbd process and the correct dcle approximation as well as the distance of each of these to the thermodynamic limit .
all of these results hold for both fixed delay and distributed delay ( see figure [ fig : schematic]a ) .
the correct dcle approximation is distinguished within the class of gaussian approximations of the dbd process by the fact that it matches both the first and second moments of the dbd process .
as we will see , it performs remarkably well at moderate system sizes in a number of dynamical settings : steady state dynamics , oscillatory dynamics , and metastable switches .
we will demonstrate via simulation and argue mathematically using characteristic functions that no other gaussian process with appropriately scaled noise performs as well . in the following
, the term ` dcle ' shall refer specifically to the dcle derived by brett and galla and expressed by , unless specifically stated otherwise .
we prove our mathematical results in the supplement @xcite .
genetic regulatory networks may be simulated using an exact dssa to account for transcriptional delay @xcite . here
we provide a heuristic derivation of a related dcle , and show that in a number of concrete examples it provides an excellent approximation of the system ( see figure [ fig : schematic ] ) .
these simulations raise the following questions : is the dcle approximation valid in general ? can the expected quality of the approximation be quantified in general ?
we answer these questions mathematically in section [ sec : main ] .
we will adopt the following notation for reactions with delay , @xmath4 here @xmath5 denotes the rate of the reaction , the dashed arrow indicates a reaction with delay , and @xmath6 is a probability measure that describes the delay distribution .
solid arrows indicate reactions without delay .
accounts for the lag between the initialization of transcription and the production of mature product . in a system with distributed delay
, different production events can have different delay times ; the order of the output process may therefore not match that of the input process .
( b e ) simulated gene regulatory network motifs : a transcriptional cascade ( b ) , oscillators ( c ) and metastable systems ( d e ) . ]
first we consider a transcriptional cascade with two genes that code for proteins @xmath7 and @xmath8 .
protein @xmath7 is produced at a basal rate ; production of @xmath8 is induced by the presence of @xmath7 .
the state of the system is represented by an ordered pair @xmath9 .
note that we use @xmath7 and @xmath8 to denote both protein names and protein numbers .
the reactions in the network , and the associated state change vectors @xmath10 , are given by @xmath11 this system can be simulated exactly using the dssa : suppose the state of the system and the reactions in the queue are known at time @xmath12 ( the queued reactions can be thought of as the `` input process '' ; see figure [ fig : schematic]a ) , and that the delay kernel @xmath13 is supported on a finite interval @xmath14 $ ] @xcite . 1 .
sample a waiting time @xmath15 from an exponential distribution with parameter @xmath16 .
2 . if there is a reaction in the queue that is scheduled to exit at time @xmath17 , advance to time @xmath18 and set @xmath19 and @xmath20 , where @xmath21 is the change in the system due to the scheduled reaction . finally ,
sample a new waiting time for the next reaction .
3 . if no reaction exits before @xmath22 , set @xmath23 and sample a reaction type from the set @xmath24 with probabilities proportional to @xmath25 , respectively . if the reaction chosen is a non - delayed reaction , perform the update @xmath26 ( or @xmath27 ) .
however , if the reaction chosen is a delayed reaction , the state change vector @xmath28 ( or @xmath29 ) is put into the queue along with an exit time @xmath18 .
the difference @xmath30 between the current and the exit time is sampled from the delay distribution @xmath13 .
we now heuristically derive the dcle for the feed - forward system from this spatially discrete process .
suppose the delay kernel @xmath6 is given by a probability density function @xmath31 supported on @xmath32 $ ] ( @xmath33 ) .
we first approximate the number of reactions that produce @xmath8 ( eq . ( [ reac : type3 ] ) ) that will be completed within the interval @xmath34 $ ] , where @xmath35 denotes the current time and @xmath36 is a small increment .
since the production of @xmath8 involves delay , a reaction of this type that is completed within @xmath34 $ ] must have been initiated at some time within @xmath37 $ ] .
let @xmath38 be the partition of @xmath37 $ ] into intervals of length @xmath36 .
the ( random ) number of reactions completed within @xmath34 $ ] _ and _ initiated within @xmath39 $ ] may be approximated by a poisson random variable with mean @xmath40 summing over @xmath41 , the ( random ) number of reactions completed within @xmath34 $ ] may be approximated by a poisson random variable with mean @xmath42;\ ] ] this is a riemann sum that approximates the integral @xmath43 known as @xmath0-leaping , this line of reasoning produces a poissonian approximation of the dbd process : @xmath44 here @xmath45 denotes a poisson random variable with mean @xmath46 . .
the inset shows sample trajectories for @xmath7 ( black ; top ) and @xmath8 ( blue ; bottom ) .
the mean field model has a fixed point , and for this reason , stochastic dynamics stay within a neighborhood of this fixed point ( see theorem [ thm : tube ] ) .
however , the effect of increasing delay is clearly seen in the cross correlation function .
the color bar displays the delay size corresponding to the cross correlation curves .
parameter values are given by @xmath47 , @xmath48 , @xmath49 , @xmath50 , @xmath51 , @xmath52 .
the cross correlations have been normalized by dividing by the standard deviations @xmath53 and @xmath54 of @xmath7 and @xmath8 .
time has been normalized to cell cycle length . ]
if these poisson random variables have large mean , they can be approximated by normal random variables .
for example , the poisson variable representing the number of reactions that produce @xmath8 can be approximated by a normal random variable with mean and variance equal to @xmath55 .
since each reaction changes the state of either @xmath7 or @xmath8 ( but never both ) , it follows that the evolution of the system can be approximated by the stochastic difference equation [ e : s - difference-1 ] @xmath56 where @xmath57 is the standard normal random variable .
system may be written in terms of concentrations .
let @xmath58 be a system size parameter .
we think of @xmath58 as a characteristic protein number ; @xmath59 and @xmath60 therefore represent fractions of this characteristic value . writing @xmath61 , @xmath62 , @xmath63 , and assuming that the basal production rate @xmath64 scales with @xmath58 as @xmath65 , we obtain [ e : s - difference-2 ] @xmath66 eq .
is the euler
maruyama type discretization of a delay stochastic differential equation .
replacing @xmath36 with @xmath67 and @xmath68 with @xmath69 in , we obtain [ e : ff - sde ] @xmath70 this is the dcle for the transcriptional cascade in this section .
taking the formal thermodynamic limit , @xmath71 , in eq .
yields the reaction rate equations derived in @xcite : [ e : ff - dde ] @xmath72 the dynamics described by eq .
are quite simple ; if @xmath73 and @xmath74 , then has a globally attracting stable stationary point .
to test the validity of the dcle approximation , we examine if it captures the interaction between the two proteins in our transcriptional cascade network .
figure [ fig : ff_cross ] shows the cross correlation functions obtained by simulating the system with @xmath75 using dssa . from left to right , the curves correspond to fixed delay @xmath0 increasing from @xmath76 to @xmath77 .
the corresponding cross correlation curves for the dcle approximation are indistinguishable from those obtained using dssa . in the heuristic derivation above , we first fix @xmath58 and let @xmath78 to obtain the dcle ; we then separately let @xmath71 to obtain the thermodynamic limit .
brett and galla @xcite also derive the dcle by first fixing @xmath58 and then sending @xmath78 .
however , the two limits , @xmath78 and @xmath71 , can not be taken independently ; this is a common problem with heuristic derivations of stochastic differential equations , even in the absence of delay @xcite .
the time discretization , @xmath79 can be thought of as a sampling frequency , while the system size , @xmath80 determines the rate at which reactions fire .
if @xmath58 becomes too large for a given @xmath81 , then the number of reactions that fire within @xmath82 $ ] no longer follows a poisson distribution with mean dependent only on the state of the system at time @xmath83 . on the other hand ,
if @xmath58 is too small for a given @xmath36 , then the poisson distribution can not be approximated by a normal distribution . in order to rigorously derive the langevin approximation and
estimate the distance between the dbd and dcle processes , we will have to take a careful limit by relating @xmath81 to @xmath58 ( with @xmath78 as @xmath71 ) .
we describe the proper scaling in section [ sec : main ] .
applied to the transcriptional cascade , theorem [ thm : dssa_dcle ] asserts that provided @xmath36 scales correctly with @xmath58 , the distance between the dbd process and the process described by eq .
converges to zero as @xmath71 ( as measured by expectations of functionals of the processes ) .
theorem [ thm : dssa_therm ] asserts that the dbd process then converges weakly to the thermodynamic limit given by eq . as @xmath71 .
moreover , when @xmath36 is correctly scaled with respect to @xmath58 , theorem [ thm : tube ] provides explicit bounds for the probabilities that the dbd and dcle processes deviate from a narrow tube around the solution of eq . .
in the previous example , the deterministic system has a fixed point .
the time series for the stochastic system , therefore , stay within a small neighborhood of this fixed point ( see inset , fig .
[ fig : ff_cross ] ) . in the next example , we show that the dcle approximation remains excellent even when the deterministic dynamics are non - trivial .
we consider a degrade - and - fire oscillator for which the deterministic system has a limit cycle .
the dcle correctly captures the peak height and the inter - peak times for the dssa realization of the degrade and fire oscillator , in addition to statistics such as the mean and variance .
the approximation does not break down at small instantaneous protein numbers .
indeed , the mathematical theory developed in this work makes an important point : protein concentrations at any particular time do not limit the quality of the dcle approximation ( in the presence of delay , or otherwise ) .
instead , the quality of the dcle approximation depends on the latent parameter @xmath58 .
theorem [ thm : tube ] makes this more precise : if one fixes the allowable error @xmath84 in the approximation of the dbd process by the dcle process , then the time @xmath85 during which the approximation error stays smaller than @xmath84 increases with @xmath58 .
, the dcle approximation given by eq
. closely matches dssa with respect to spike height distribution ( b ) and interspike interval distribution ( c ) .
in contrast , removing delay from the diffusion term results in a poor approximation of these distributions , as shown by the sizable shifts affecting the red curves .
( d ) and ( e ) illustrate mean repressor protein level and repressor protein variance , respectively , as functions of system size @xmath58 .
. provides a good approximation for all simulated values of @xmath58 while the performance of eq .
improves as @xmath58 increases .
the quantity @xmath86 represents protein number , not protein concentration .
parameter values are @xmath87 , @xmath88 , @xmath89 , @xmath90 , @xmath91 , @xmath92 .
a soft boundary was added at @xmath76 to ensure positivity . ]
the degrade and fire oscillator depicted schematically in figure [ fig : schematic]c consists of a single autorepressive gene and corresponds to the reaction network @xmath93 the production rate @xmath94 is given by @xmath95 , where @xmath96 is the propensity function @xmath97 the enzymatic degradation rate @xmath98 is given by @xmath99 .
here @xmath100 ; @xmath1 is the michaelis - menten constant , @xmath101 the maximal enzymatic degradation rate , and @xmath102 the dilution rate coefficient . in the thermodynamic limit
, the system is modeled by the delay differential equation @xmath103 as before , @xmath104 denotes the concentration of protein @xmath7 .
we model the formation of functional repressor protein using distributed delay ( described by the probability measure @xmath6 ) ; this delayed negative feedback can induce oscillations @xcite .
figure [ fig : df_oscillator]a depicts a sample realization of the stochastic version of the degrade and fire oscillator ( the finite system size regime ) generated by dssa .
the dcle approximation is in this case given by @xmath105^{\frac{1}{2 } } { \mathrm{d } } w_{t}. \end{aligned}\ ] ] figure [ fig : df_oscillator ] illustrates that eq . provides a good approximation of the dbd dynamics , even when system size is relatively small . at system size @xmath106 ,
the spike height distribution and interspike interval distribution obtained using the dssa ( black dots in figure [ fig : df_oscillator]b[fig : df_oscillator]c ) are nearly indistinguishable from those obtained using eq .
( black curves in figure [ fig : df_oscillator]b[fig : df_oscillator]c ) .
further , we see a close match with respect to mean repressor protein level and repressor protein variance across a range of system sizes ( figure [ fig : df_oscillator]d[fig : df_oscillator]e ) .
interestingly , the dcle approximation is very good even though the protein number approaches zero during part of the oscillation .
this illustrates a central feature of the theory : the quality of the dcle approximation is a function of a latent parameter @xmath58 , not of the number of molecules present at any given time .
the exact form of the diffusion term is crucial to the accuracy of the dcle approximation . if we remove delay from the diffusion term in eq .
, we obtain @xmath107^{1/2 } { \mathrm{d } } w_{t}. \end{aligned}\ ] ] at system size @xmath106 ( red curves in figure [ fig : df_oscillator]b[fig : df_oscillator]c ) , dsde produces dramatically different results from those generated by the correct dcle approximation . the performance of eq
. improves as @xmath58 increases ( figure [ fig : df_oscillator]d[fig : df_oscillator]e ) .
this is expected , as both eq . and
eq . converge weakly to eq . as @xmath108 .
understanding metastability in stochastic systems is of fundamental importance in the study of biological switches @xcite .
while metastability is well understood mathematically in the absence of delay , understanding the impact of delay on metastability remains a major theoretical and computational challenge @xcite .
we examine two canonical examples to show that the dcle can be used to study the impact of delay on metastability : a positive feedback circuit and a co - repressive genetic toggle switch . and a hill coefficient of @xmath109 .
the bottom row corresponds to a delay @xmath110 and @xmath111 .
the tail of the hitting times distribution becomes longer with both increasing delay and increasing hill coefficient @xmath112 .
the dcle captures the lengthening of the tail due to both effects .
parameter values are @xmath113 , @xmath114 , @xmath115 , @xmath116 , @xmath52 , @xmath75 . ]
the simplest metastable system consists of a single protein that drives its own production ( figure [ fig : schematic]d ) .
the chemical reaction network is given by @xmath117 with @xmath118 for the propensity @xmath119 in the thermodynamic limit , the dynamics of this model are described by the dde @xmath120 here @xmath121 represents protein concentration and @xmath112 is the hill coefficient . in the thermodynamic limit , there are two stable stationary states , @xmath122 and @xmath123 , as well as an unstable stationary state @xmath124 .
these states satisfy @xmath125 .
in the stochastic ( finite @xmath58 ) regime , the stationary states @xmath122 and @xmath123 become metastable .
we simulate the metastable dynamics using dssa and the dcle approximation given in this case by @xmath126^{\frac{1}{2 } } \mathrm{d } w_{t}. \end{aligned}\ ] ] figure [ fig : pf_hitting ] displays hitting time distributions for the dssa simulations ( black curves ) and eq .
( blue curves ) . a hitting time is defined as follows : we choose neighborhoods @xmath127 and @xmath128 of @xmath122 and @xmath124 , respectively .
we start the clock when a trajectory enters @xmath127 from the right .
the clock is stopped when that trajectory first enters @xmath128 .
a hitting time is the amount of time that elapses from clock start to clock stop .
we see that for no delay ( figure [ fig : pf_hitting ] , top ) and fixed delay @xmath129 ( bottom ) , the dcle approximation accurately captures the hitting time distributions for hill coefficients increasing from @xmath130 to @xmath131 .
hence , the dcle approximation accurately captures the rare events associated with a spatially - discrete delay stochastic process .
this is significant because dsdes are more amenable to large deviations theoretical analysis than their spatially - discrete counterparts . hitting times increase dramatically as the delay increases from @xmath76 to @xmath132 , in accord with earlier analysis @xcite .
a dramatic increase is also seen as the hill coefficient increases .
this is due to the fact that the potential wells around @xmath122 and @xmath123 deepen as @xmath112 increases . and fall back into the same neighborhood before transitioning into a neighborhood of the stable point @xmath133 .
the two right panels illustrate trajectories that leave a neighborhood of @xmath134 and transition to a neighborhood of @xmath133 before falling back into the first neighborhood .
top panels correspond to dssa ; bottom panels correspond to dcle .
cartoons of typical trajectories corresponding to failed transitions ( left ) and successful transitions ( right ) are shown for the dcle process .
plots are shown for @xmath135 .
the values of the other parameters are @xmath136 , @xmath137 , @xmath116 . ] the co - repressive toggle switch ( figure [ fig : schematic]e ) is a two - dimensional metastable system described in the thermodynamic limit by the ddes [ e : toggle_dde ] @xmath138 the measure @xmath139 describes the delay associated with production in this symmetric circuit .
. has two stable stationary points @xmath140 and @xmath141 separated by the unstable manifold associated with a saddle equilibrium point @xmath142 . in the stochastic ( finite system size ) regime ,
the stable stationary points become metastable . in this regime
a typical trajectory spends most of its time near the metastable points , occasionally moving between them .
figure [ fig : toggle ] displays density plots corresponding to trajectories that either successfully transition between metastable states ( four panels on the right ) or make failed transition attempts ( four panels on the left ) . even for the moderate system size @xmath135 , the density plots generated by dssa ( top four panels ) closely match those generated by the dcle approximation in eq .
( bottom four panels ) . given the importance of rare events throughout stochastic dynamics , it is encouraging that the dcle approximation captures their statistics well .
the simulations thus far described suggest that the dcle closely approximates the dbd process provided that @xmath36 scales properly with @xmath58 .
we next provide mathematical statements that make this observation precise .
we prove that the distance between the dbd process and the approximating dcle ( as measured by expectations of functionals of the processes ) converges to zero as the system size @xmath71 ( theorem [ thm : dssa_dcle ] ) .
in particular , theorem [ thm : dssa_dcle ] implies that the dcle may be used to approximate all moments of the dbd process .
further , we then prove that the dbd and dcle processes both converge weakly to the thermodynamic limit ( theorem [ thm : dssa_therm ] ) .
theorem [ thm : dssa_therm ] strengthens a result of schlicht and winkler @xcite that establishes convergence of the first moment of the dbd process .
theorem [ thm : tube ] quantitatively bounds the probabilities that the dbd and dcle processes deviate from a narrow tube around the solution of the deterministic thermodynamic limit .
we first precisely describe the general setting and then state our theorems .
all proofs are provided in the supplement @xcite .
consider a system of @xmath3 biochemical species and @xmath143 possible reactions .
we are interested in describing the dynamics as a function of a latent system parameter @xmath58 , the system size .
let @xmath144 denote the state of the system at time @xmath83 .
each reaction @xmath145 is described by the following : 1 . a propensity function @xmath146 .
the firing rate of reaction @xmath145 is given by @xmath147 .
2 . a state - change vector @xmath148 .
the vector @xmath149 describes the change in the number of molecules of each species that results from the completion of a reaction of type @xmath150 .
3 . a probability measure @xmath151 supported on @xmath152 $ ] .
the measure @xmath151 models the delay that may occur between the initiation and completion of a reaction of type @xmath150 .
if reaction @xmath145 is instantaneous , then @xmath153 .
if the delay is a fixed value @xmath154 , then @xmath155 . if @xmath151 has a density , we denote it by @xmath156 . given a system trajectory up to time @xmath83 , @xmath157
, the dssa can be described as follows : 1 .
[ i : time_reaction ] sample a time @xmath158 to the next reaction from the exponential distribution with rate @xmath159 2 . select a reaction @xmath160 with probability @xmath161 3 .
sample a delay time @xmath0 from @xmath162 .
if no delayed reactions are set to finish in the time interval @xmath163 $ ] , then proceed as follows . if @xmath164 , then move to time @xmath165 and set @xmath166 .
if @xmath154 , then move to time @xmath165 , set @xmath167 , and put the state change vector @xmath168 into a queue along with the designated time of reaction completion , @xmath169 .
if a reaction from the past is set to complete in @xmath163 $ ] , then move to the time @xmath170 of the completion of the first such reaction , update @xmath171 accordingly , and then proceed to ( [ i : time_reaction ] ) .
schlicht and winkler @xcite prove the existence of the stochastic process @xmath172 from which the dssa samples .
let @xmath173 denote the probability measure on realizations @xmath174 associated with the scaled process @xmath175 .
these realizations lie in the space @xmath176 of right - continuous functions from @xmath177 $ ] into @xmath178 that possess limits from the left .
our main results quantify the behavior of @xmath173 as @xmath71 .
we make the following regularity assumptions on the propensities @xmath179 . 1 .
[ i : propen_reg ] the functions @xmath179 have continuous derivatives of order @xmath180 .
[ i : propen_supp1 ] there exists a compact set @xmath1 such that @xmath181 for all @xmath182 .
[ i : propen_supp2 ] for all @xmath182 , @xmath183 only if all coordinates of the vector @xmath184 are nonnegative ; @xmath185 otherwise .
the correct dcle approximation of @xmath175 is given for @xmath186 by @xmath187 where @xmath188 is a @xmath3-dimensional vector of independent standard brownian motions and @xmath189 is given by @xmath190 let @xmath191 denote the stochastic process described by eq . and let @xmath192 denote the probability measure on realizations @xmath193 associated with this process .
theorem [ thm : dssa_dcle ] controls the distance between @xmath173 and @xmath192 .
[ thm : dssa_dcle ] assume that the propensities @xmath179 satisfy ( [ i : propen_reg])([i : propen_supp2 ] ) .
fix @xmath194 .
for every continuous observable @xmath195 , we have @xmath196 as @xmath71 .
theorem [ thm : dssa_therm ] establishes weak convergence of the scaled process @xmath175 to the thermodynamic limit governed by the delay reaction rate equations @xmath197 [ thm : dssa_therm ] assume that the propensities @xmath179 satisfy ( [ i : propen_reg])([i : propen_supp2 ] ) .
fix @xmath194 .
for every continuous observable @xmath195 , we have @xmath198 as @xmath71 , where @xmath199 denotes the solution of eq . .
crucial to the proofs of theorems [ thm : dssa_dcle ] and [ thm : dssa_therm ] are the discretization of the time interval @xmath177 $ ] and the development of quantitative controls on processes that approximate @xmath200 .
we partition @xmath177 $ ] into subintervals of length @xmath201 .
it is crucial that @xmath202 scales with @xmath58 in the right way .
the quantitative pathwise controls in theorem [ thm : tube ] hold if @xmath203 .
we define an euler - maruyama discretization @xmath204 of . for @xmath205 ,
define @xmath206 . for integers @xmath207 and @xmath208
, define @xmath209 recursively by @xmath210 where @xmath211 \delta , \\ a_{2 } & = \frac{\sqrt{\delta}}{\sqrt{n}}\eta , \end{aligned}\ ] ] and @xmath212 is a mean @xmath76 multivariate gaussian random variable with correlation matrix @xmath213 defined as @xmath214 for @xmath215 , define @xmath216 by linearly interpolating between @xmath217 and @xmath218 .
theorem [ thm : tube ] asserts that realizations of @xmath219 stay close to those of @xmath204 with high probability .
[ thm : tube ] suppose that @xmath203 .
there exist constants @xmath220 and @xmath221 such that @xmath222 where @xmath223 is a system constant defined by @xmath224 and @xmath225 is the ` discretized ' norm @xmath226
stochastic differential equations ( sdes ) are one of our main tools for modeling noisy processes in nature . interactions between the components of a system or network are frequently not instantaneous .
it is therefore natural to include such delay into corresponding stochastic models .
however , the relationship between delay sdes ( dsdes ) and the processes they model has not been fully established .
delay stochastic differential equations have previously been formally derived from the delay chemical master equation ( dcme ) . unlike the chemical master equation , however , the dcme is not closed ; this complicates the derivation of dsde approximations .
closure in this context means the following : let @xmath227 denote the probability that the stochastic system is in state @xmath228 at time @xmath83 .
the dcme expresses the time derivative of @xmath227 in terms of joint probabilities of the form @xmath229 - the probability that the system is in state @xmath150 at time @xmath83 and was in state @xmath230 at time @xmath231 , where @xmath0 is the delay .
the one - point probability distribution @xmath232 is therefore expressed in terms of two - point joint distributions , resulting in a system that is not closed .
timescale separation assumptions have been used to close the dcme .
if the delay time is large compared to the other timescales in the system , one may assume that events that occur at time @xmath233 are decoupled from those that occur at time @xmath83 and close the dcme @xcite by assuming the joint probabilities may be written as products : @xmath234 having closed the dcme , one may then derive dsde approximations @xcite as well as useful expressions for autocorrelations and power spectra @xcite .
approximations of dsde type have also been derived using system size expansions such as van kampen expansions and kramers - moyal expansions for both fixed delay @xcite and distributed delay @xcite .
brett and galla @xcite use the path integral formalism of martin , siggia , rose , janssen , and de dominicis to derive the delay chemical langevin equation ( dcle ) without relying on the dcme . using this formalism , a moment generating functional
may be expressed in terms of the system size parameter @xmath58 and the sampling rate @xmath81 . in the continuous - time limit , @xmath78
, the dcle may be inferred from the moment generating functional .
however , the brett and galla derivation has some limitations .
first , the @xmath78 limit can not be taken without simultaneously letting @xmath108 .
intuitively , this is because as @xmath78 , the gaussian approximation to the poisson distribution with mean @xmath235 breaks down unless the parameter @xmath236 simultaneously diverges to infinity .
second , the derivation gives no quantitative information about the distance between the dcle and the original delay birth - death ( dbd ) process . in this paper , we address these shortcomings .
we prove rigorously that the dbd process can be approximated by a class of gaussian processes that includes the dcle .
in particular , we establish that for most biophysically relevant propensity functions , the dcle process will approximate all moments of the dbd process .
the rigorous proof includes bounds on the quality of the approximation in terms of the time @xmath85 for which the approximation is desired to hold and the characteristic protein number @xmath58 ( see theorem [ thm : tube ] ) .
the error bounds also indicate that the quality of the dcle approximation worsens with increasing upper bounds on the reaction propensity functions and state - change vectors .
physically , this means that high reaction rates and reactions that cause large changes in the protein populations are detrimental to the quality of the dcle approximation .
the dcle is one of many gaussian processes that approximate the dbd process . among all gaussian approximations with noise components that scale as @xmath237 ,
the dcle is optimal because it is the only such approximation that exactly matches the first and second moments of the dbd process .
we formally justify this assertion in the supplement @xcite using characteristic functions .
as our simulations of the degrade and fire oscillator demonstrate , the dcle can significantly outperform other gaussian approximations at moderate system sizes .
nevertheless , the quantitative tube estimates in theorem [ thm : tube ] apply to _ any _ gaussian approximation of the dbd process provided the noise scales as @xmath237 .
this is significant because it is often advantageous to use linear noise approximations of the dcle .
delay appears in the drift component of a linear noise approximation but not in the diffusion component .
linear noise approximations are therefore easier to analyze than their dcle counterparts . in particular ,
elements of the theory of large deviations for markovian systems can be extended to sdes with delay in the drift @xcite . for metastable systems ,
our simulations indicate that the dcle captures both temporal information ( such as hitting times for the positive feedback model ; see fig .
[ fig : pf_hitting ] ) and spatial information ( such as densities for trajectories corresponding to failed and successful transitions ; see fig . [
fig : toggle ] ) .
this suggests that dcle approximations may be used to study rare events for biochemical systems that exhibit metastability .
we have shown that the dcle provides an accurate approximation of a number of stochastic processes . although we chose gene regulatory networks in our examples ,
the theory is applicable to general birth - death processes with delayed events .
sdes , and the chemical langevin equation in particular , are fundamental in modeling and understanding the behavior of natural and engineered systems .
we therefore expect that the dcle will be widely applicable when delays impact system dynamics .
consider a system of @xmath3 biochemical species and @xmath143 possible reactions .
we are interested in describing the dynamics as a function of a latent system parameter @xmath58 , the system size .
let @xmath144 denote the state of the system at time @xmath83 .
each reaction @xmath145 is described by the following : 1
. a propensity function @xmath146 .
the firing rate of reaction @xmath145 is given by @xmath147 .
2 . a state - change vector @xmath148 .
the vector @xmath149 describes the change in the number of molecules of each species that results from the completion of a reaction of type @xmath150 .
3 . a probability measure @xmath151 supported on @xmath152 $ ] .
the measure @xmath151 models the delay that may occur between the initiation and completion of a reaction of type @xmath150
. if reaction @xmath145 is instantaneous , then @xmath153 . if the delay is a fixed value @xmath154 , then @xmath155 . if @xmath151 has a density , we denote it by @xmath156 . given a system trajectory up to time @xmath83 , @xmath157
, the dssa can be described as follows : 1 .
[ i : time_reaction ] sample a time @xmath158 to the next reaction from the exponential distribution with rate @xmath159 2 .
select a reaction @xmath160 with probability @xmath161 3 .
sample a delay time @xmath0 from @xmath162 .
if no delayed reactions are set to finish in the time interval @xmath163 $ ] , then proceed as follows . if @xmath164 , then move to time @xmath165 and set @xmath166 .
if @xmath154 , then move to time @xmath165 , set @xmath167 , and put the state change vector @xmath168 into a queue along with the designated time of reaction completion , @xmath169 .
if a reaction from the past is set to complete in @xmath163 $ ] , then move to the time @xmath170 of the completion of the first such reaction , update @xmath171 accordingly , and then proceed to ( [ i : time_reaction ] ) .
schlicht and winkler @xcite prove the existence of the stochastic process @xmath172 from which the dssa samples .
let @xmath173 denote the probability measure on realizations @xmath174 associated with the scaled process @xmath175 .
these realizations lie in the space @xmath176 of right - continuous functions from @xmath177 $ ] into @xmath178 that possess limits from the left .
our main results quantify the behavior of @xmath173 as @xmath71 .
we make the following regularity assumptions on the propensities @xmath179 . 1 .
[ i : propen_reg ] the functions @xmath179 have continuous derivatives of order @xmath180 .
[ i : propen_supp1 ] there exists a compact set @xmath1 such that @xmath181 for all @xmath182 .
[ i : propen_supp2 ] for all @xmath182 , @xmath183 only if all coordinates of the vector @xmath184 are nonnegative ; @xmath185 otherwise .
the correct dcle approximation of @xmath175 is given for @xmath186 by @xmath238 where @xmath188 is a @xmath3-dimensional vector of independent standard brownian motions and @xmath189 is given by @xmath190 let @xmath191 denote the stochastic process described by eq . and let @xmath192 denote the probability measure on realizations @xmath193 associated with this process .
theorem [ thm : dssa_dcle ] controls the distance between @xmath173 and @xmath192 .
[ thm : dssa_dcle ] assume that the propensities @xmath179 satisfy ( [ i : propen_reg])([i : propen_supp2 ] ) .
fix @xmath194 . for every continuous observable @xmath195
, we have @xmath196 as @xmath71 .
theorem [ thm : dssa_therm ] establishes weak convergence of the scaled process @xmath175 to the thermodynamic limit governed by the delay reaction rate equations @xmath197 [ thm : dssa_therm ] assume that the propensities @xmath179 satisfy ( [ i : propen_reg])([i : propen_supp2 ] ) .
fix @xmath194 .
for every continuous observable @xmath195 , we have @xmath198 as @xmath71 , where @xmath199 denotes the solution of eq . .
crucial to the proofs of theorems [ thm : dssa_dcle ] and [ thm : dssa_therm ] are the discretization of the time interval @xmath177 $ ] and the development of quantitative controls on processes that approximate @xmath200 .
we partition @xmath177 $ ] into subintervals of length @xmath201 .
it is crucial that @xmath202 scales with @xmath58 in the right way .
the quantitative pathwise controls in theorem [ thm : tube ] hold if @xmath203 .
we define an euler - maruyama discretization @xmath204 of . for @xmath205 ,
define @xmath206 . for integers @xmath207 and @xmath208 , define @xmath209 recursively by @xmath239 \delta + \frac{\sqrt{\delta}}{\sqrt{n}}\eta , \ ] ] where @xmath212 is a mean @xmath76 multivariate gaussian random variable with correlation matrix @xmath213 defined as @xmath214 for @xmath215 , define @xmath216 by linearly interpolating between @xmath217 and @xmath218 . theorem [ thm : tube ] asserts that realizations of @xmath219 stay close to those of @xmath204 with high probability .
[ thm : tube ] suppose that @xmath203 .
there exist constants @xmath220 and @xmath221 such that @xmath222 where @xmath223 is a system constant defined by @xmath224 and @xmath225 is the ` discretized ' norm @xmath226
theorem [ thm : dssa_therm ] establishes weak convergence of the scaled process @xmath240 to the thermodynamic limit .
we prove theorem [ thm : dssa_therm ] in two steps .
first , we show that the family of measures @xmath241 is a tight family on @xmath176 . by the prohorov theorem ( see _ e.g. _
@xcite ) , the family @xmath241 is then relatively compact in the space of probability measures .
second , we consider finite sequences of times @xmath242 in @xmath177 $ ] and study the finite - dimensional distributions associated with @xmath243 in order to show that the family @xmath241 has a weak limit and to characterize this limit . * scaling of the time discretization * @xmath244*. * we assume throughout that @xmath201 satisfies @xmath245 , where @xmath246 and @xmath247 are constants and @xmath248 . intuitively , @xmath36 must be sufficiently small so that propensity functions do not change significantly over any time interval of length @xmath36 and sufficiently large so that many reactions fire over any such interval .
the space @xmath176 is a metric space with skorohod metric @xmath249 defined as follows .
let @xmath250 denote the set of nondecreasing functions @xmath251 \to [ 0,t]$ ] with @xmath252 and @xmath253 . for @xmath254
, define @xmath255 for functions @xmath256 and @xmath257 in @xmath176 , define @xmath258 } { \left\lvert z_{1 } ( t ) - z_{2 } ( \gamma ( t ) ) \right\rvert } { \leqslant}{\varepsilon}\big\}.\end{aligned}\ ] ] let @xmath259 be a sequence of probability measures on @xmath176 .
we recall a characterization of tightness for @xmath259 . for @xmath260 and @xmath261 , define @xmath262 where the infimum is taken over partitions of @xmath177 $ ] such that @xmath263 for all @xmath41 . 1 .
[ i : tight1 ] for every @xmath264 , there exists @xmath64 such that @xmath265 for all @xmath266 .
[ i : tight2 ] for all @xmath267 and @xmath264 , there exist @xmath261 and @xmath268 such that @xmath269 for all @xmath270 .
it follows from the markov inequality that @xmath274 and so it suffices to minimize the function @xmath275 it is easy to see that the minimum is reached at @xmath276 , provided @xmath272 . on substituting back into @xmath277
we get the result . 1 .
suppose @xmath278 .
there exists a constant @xmath279 such that @xmath280 \right)}{\left(\frac{x}{y}\right)^{nx } } \leq e^{-c_3 ny } \to 0 \quad \text { as } n\to \infty\ ] ] 2 .
let @xmath281 , @xmath282 and @xmath283 .
let @xmath284 there exists a constant @xmath285 such that @xmath280 \right)}{\left(\frac{x}{y}\right)^{nx } } \leq e^{-c_4 nx } \to 0 \quad \text { as } x\to \infty\ ] ] first inequality : since @xmath286 , the expression above is an indeterminate form of the type @xmath287 . on computing logarithms
, we need to show that the expression below diverges to @xmath288 : @xmath289\ ] ] this happens if and only if @xmath290 for @xmath291 , which , in turn , is always true when @xmath292 .
the proof of the second inequality is similar .
[ l : firing_bound ] let @xmath296 be the event that reaction @xmath297 fires at least @xmath298 times on @xmath34 $ ] . using lemma [ lemma : a1 ] and lemma [ lemma : a2 ] , we have @xmath299\right)}{\left ( \frac{2 \cdot \max{\ { |f_i|_\infty \}}}{|f_{i_{0}}|_\infty}\right)^{2n \delta \cdot \max{\ { |f_i|_\infty \ } } } } \leq e^{-2 c_4 n \delta \max_{i } |f_i|_\infty } { \mathrel{=\mkern-4.2mu\raise.095ex\hbox{$:$}}}e^{-c_5 n^{1-\alpha}}.\ ] ] [ dcle : supp : exponentials_are_close ] for every constant @xmath302 , for @xmath303 large enough , @xmath304 ~d\mu_j(s ) \right ] \right ) \right.\\ - \left . \exp\left ( n\delta \left [ \sum_{j=1}^m \left ( e^{\mathrm{i}\mathbf{r}v_j / n}-1\right ) \int_{0}^{\infty } [ f_j(b_n(t -s ) ) + c\delta\zeta ]
~d\mu_j(s ) \right ] \right ) \right|\\ \hfill \leq 2c \delta^2 \zeta |\mathbf{r}| \left(\sum_{j = 1}^m |v_j|\right ) + \frac{c}{2 } |\mathbf{r}|^2 \delta^3 \zeta \left ( \sup_{j } m |f_j|_\infty |v_j|\right)^2 . \end{aligned}\ ] ] @xmath304 ~d\mu_j(s ) \right ] \right ) \right.\\ - \left .
\exp\left ( n\delta \left [ \sum_{j=1}^m \left ( e^{\mathrm{i}\mathbf{r}v_j / n}-1\right ) \int_{0}^{\infty } [ f_j(b_n(t -s ) ) + c\delta\zeta ] ~d\mu_j(s ) \right ] \right ) \right| \\
\leq \left| \exp\left ( n\delta \sum_{j = 1}^m \left\{\int_{0}^{\infty } \left[f_j(b_n(t -s ) ) - c\delta \zeta\right]~ d\mu_j(s)\right\ } \left ( \frac{\mathrm{i } \mathbf{r } v_j}{n } - \frac{(\mathbf{r}v_j)^2}{n^2 } + \mathcal{o}(n^{-3})\right ) \right)\right.\\ \left . - \exp\left ( n\delta \sum_{j = 1}^m \left\{\int_{0}^{\infty } \left[f_j(b_n(t -s ) ) + c\delta \zeta\right]~ d\mu_j(s)\right\ } \left ( \frac{\mathrm{i } \mathbf{r } v_j}{n } - \frac{(\mathbf{r}v_j)^2}{n^2 } + \mathcal{o}(n^{-3})\right ) \right ) \right|\\ \leq \left| n\delta \sum_{j = 1}^m
\left ( \frac{\mathrm{i}\mathbf{r}v_j}{n } - \frac{(\mathbf{r}v_j)^2}{n^2 } + \mathcal{o}(n^{-3})\right ) \left ( \int_{0}^{\infty } 2 c\delta \zeta ~ d\mu_j(s)\right ) \right|\\ + \frac{1}{2 } \left| n^2\delta^2 \left\ { \sum_{j = 1}^m \left ( \frac{\mathrm{i}\mathbf{r}v_j}{n } - \frac{(\mathbf{r}v_j)^2}{n^2 } + \mathcal{o}(n^{-3})\right)^2\right.\right.\\ \left.\left . \left\
{ \int_{0}^\infty \left[(f_j(b_n(t - s ) ) - c \delta \zeta)^2 -(f_j(b_n(t - s ) ) + c \delta \zeta)^2 \right ] d\mu_j(s ) \right\ } \right\ } \right|\\ \leq 2c \delta^2 \zeta |\mathbf{r}| \left ( \sum_{j = 1}^{m } |v_j|\right ) + \frac{c}{2}|\mathbf{r}|^2 \delta^3 \zeta \left ( \sum_{j = 1}^m \int_{0}^\infty |v_j f_j(t - s)|~ d\mu_j(s ) \right)^2 \end{aligned}\phantom{\hfill}\ ] ] [ dcle : prop:3 ] define @xmath305 . for @xmath306 and @xmath282 , we have @xmath307 - e\left [ \exp\left ( n \delta \sum_{j = 1}^m\left ( e^{\mathrm{i}\mathbf{r } v_j /n } -1\right ) \int_{0}^{\infty } f_j ( b_n(t_0 -s ) ) ~d\mu(s ) \right ) \right ] \right|\\ \leq 5 |\mathbf{r}| \delta^2 \zeta |f'|_\infty + o(\delta^{2+\epsilon } ) \end{aligned}\ ] ] where @xmath308 since @xmath309 , we may also take @xmath310 corollary [ c : c_jump_bound ] gives @xmath311 .
define a random variable @xmath312 as follows : @xmath313 abusing notation , we write @xmath314 for @xmath312 throughout the proof .
one can see that @xmath315 pointwise , and by the dominated convergence theorem we can directly obtain that @xmath316 \to e[\varphi]$ ] .
however , in doing so we do nt obtain a rate .
for this reason , we perform a more careful computation .
define sets @xmath317 as the set of all outcomes for which @xmath318 for all @xmath319 .
the complement of this set is denoted as @xmath320 .
if for any @xmath321 , @xmath322 , and so from equation it follows that @xmath323 ( by lemma [ lemma : a2 ] ) .
it then follows that @xmath324 and so @xmath325 - e[\varphi_{n\delta\zeta}]| \leq \zeta\sum_{l \geq 1 } ( l+1)n\delta e^{-c_4l n \delta\zeta}.$ ] the bound on the right is smaller than @xmath326 an analogous computation can be used to show that @xmath327 - e[e^{\mathrm{i}\mathbf{r } \varphi_{n\delta\zeta}/n}]| \leq 4 \zeta e^{-c_4 \zeta n^{1 - \alpha}}.$ ] therefore , it is enough to approximate @xmath328 $ ] for @xmath329 .
let @xmath330 be the random variable that defines the change to a process over the interval @xmath331 $ ] that has constant propensity functions @xmath332 for the reaction that contributes a change @xmath333 to the system .
call @xmath334 the stationary measure for such a process .
it follows from definition that @xmath335 - e[e^{\mathrm{i}\mathbf{r } \tilde{\varphi}/n } ] \right| \leq \bigg| \int_{\varphi \leq k } e^{\mathrm{i}\mathbf{r } \varphi_k / n}~ d\mathbb{b}_n - & \int_{\varphi \leq k } e^{\mathrm{i}\mathbf{r}\tilde{\varphi}/n}~d\tilde{p_n } \bigg| \\ & { } + \sum_{l \geq k+1 } \tilde{p_n}(l ) + \sum_{l \geq k+1 } \mathbb{b}_n(l ) .
\end{split}\ ] ] @xmath336 is abbreviation for @xmath337 ( and analogously for @xmath338 ) .
we now compute bounds for the first term on the right .
recall that @xmath339 .
since the largest size of any co - ordinate of @xmath330 is smaller than @xmath340 , the arguments to the propensity functions change by at most @xmath341 . from the smoothness of the propensity functions it follows that for any @xmath342 , @xmath343 the estimate in the middle is the infinitesimal rate for the process @xmath344 , and can be approximated only in terms of information available up to time @xmath12 .
therefore , @xmath345 \\ & { } - e\left [ e^{n\delta \sum_{j = 1}^m(e^{\mathrm{i}\mathbf{r}v_j / n } -1)(\int_{0}^{\infty } f_j(b_n(t_0 -s ) ) ~ d\mu_j(s ) + \delta \zeta |f'|_\infty ) } \right ] \big| \\ & \leq 5 |f'|_\infty \delta^2 \zeta |\mathbf{r}| \left(\sum_{j = 1}^m |v_j|\right ) + \mathcal{o}(\delta^{2+\epsilon } ) \end{aligned}\ ] ] by lemma [ dcle : supp : exponentials_are_close ] .
the other terms have tails of the order of @xmath346 and so , for large enough @xmath58 , the dominant term is the one involving @xmath347 .
finally , we observe that @xmath348 = e\left [ \exp\left ( n \delta \sum_{j = 1}^m\left ( e^{\mathrm{i}\mathbf{r } v_j /n } -1\right ) \int_{0}^{\infty } f_j ( b_n(t_0 -s ) ) ~d\mu(s ) \right ) \right].\ ] ] consider the approximation to the characteristic function for the jump @xmath349 on the interval @xmath350 $ ] . on expanding the term @xmath351 term using a taylor series expansion , we can approximate the expression @xmath352\ ] ] by @xmath353\ ] ] with an error that is @xmath354 .
however , this is the characteristic function for the gaussian random variable @xmath355 \delta + \frac{\sqrt{\delta}}{\sqrt{n } } \eta \end{aligned}\]][dcle : supp : e : sigma ] where @xmath212 is a mean @xmath76 multivariate gaussian random variable with correlation matrix @xmath213 , with @xmath356 [ dcle : supp : defn : dcle ] for @xmath357 , define @xmath206 . for @xmath358 , define @xmath209 recursively as @xmath359 \delta + \frac{\sqrt{\delta}}{\sqrt{n}}\eta\ ] ] where @xmath212 is a mean 0 multivariate gaussian random variable with correlation matrix @xmath213 defined as @xmath360 the next proposition estimates the probability of finding the processes @xmath219 and @xmath365 outside a tube around @xmath366 of radius greater than @xmath367 . as the proposition shows , the gaussian process has a smaller tail ( of the order @xmath368 ) than the birth - death process ( which has a tail of order @xmath369 ) .
let @xmath349 denote the increment to the process @xmath373 in the time interval @xmath374 $ ] . from lemma [ lemma :
a1 ] it follows that for @xmath58 large enough @xmath375\delta \right\| > \frac{\zeta}{\sqrt{n\delta } } \right ) \leq \delta^{\delta^{-1}}.\ ] ] let @xmath376 denote the increment to the process @xmath366 on @xmath377.$ ] since @xmath378 for @xmath357 , it follows that @xmath379 we now bound the error on the interval @xmath380.$ ] @xmath381 except on a set of measure @xmath382 the bound on the first term is @xmath383 and the bound on the middle term is @xmath384 ; the total error is smaller than @xmath385 this argument can now be propagated forward . for any finite @xmath230 , except on a set of @xmath386 measure smaller than @xmath387 , the distance between @xmath388 and @xmath389 is at most @xmath390 the summation on the right can be bounded by @xmath391 . since @xmath230 is at most @xmath392 , the contribution of the summation is small relative to the contribution of the first term .
therefore , we can bound the distance between realizations by @xmath393 .
the proof for the process @xmath365 proceeds analogously . for @xmath58 large enough @xmath394\delta \right\| >
\frac{\zeta}{\sqrt{n\delta } } \right ) \leq e^{-\delta^{-2}/2}\ ] ] from where it follows that @xmath395 this bound can then be propagated as before . 51ifxundefined [ 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 10.1111/j.1365 - 2958.2010.07111.x
[ * * , ( ) ] link:\doibase 10.1073/pnas.0503858102 [ * * , ( ) ] , \doibase http://dx.doi.org/10.1016/0065-2571(65)90067-1 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/s0960-9822(03)00534-7 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.102.068105 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/s0960-9822(03)00494-9 [ * * , ( ) ] @noop * * , ( ) link:\doibase 10.1038/nature07389 [ * * , ( ) ] link:\doibase 10.1038/nature07616 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/s0006-3495(02)75249-1 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/j.jtbi.2004.04.006 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/j.cell.2010.12.019 [ * * , ( ) ] link:\doibase 10.1073/pnas.0913317107 [ * * , ( ) ] link:\doibase 10.1038/ncomms1422 [ * * ( ) , 10.1038/ncomms1422 ] link:\doibase 10.1371/journal.pone.0002972 [ * * , ( ) ] link:\doibase 10.1103/physreve.80.031129 [ * * , ( ) ] @noop ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1038/nature02298 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/s0006-3495(01)75949-8 [ * * , ( ) ] @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) link:\doibase 10.1103/physrevlett.111.058104 [ * * , ( ) ] link:\doibase 10.1142/s0219493705001389 [ * * , ( ) ] link:\doibase 10.1080/07362990500397715 [ * * , ( ) ] link:\doibase 10.1371/journal.pcbi.1002264 [ * * , ( ) ] http://www.scopus.com/inward/record.url?eid=2-s2.0-33645429016&partnerid=40&md5=47e48b2d9e17bcc889478b564877d5e0 [ * * , ( ) ] link:\doibase 10.1137/060666457 [ * * , ( ) ] http://www.scopus.com/inward/record.url?eid=2-s2.0-31544465969&partnerid=40&md5=2871efe61543fb533e53de65f2d40e04 [ * * ( ) ] http://www.scopus.com/inward/record.url?eid=2-s2.0-34250754425&partnerid=40&md5=143efd47f6b1d958a31e5de896e1bd59 [ * * ( ) ] http://www.scopus.com/inward/record.url?eid=2-s2.0-0035933994&partnerid=40&md5=0cbe82982da7f4d67dd3f6288097f278 [ * * , ( ) ] http://www.scopus.com/inward/record.url?eid=2-s2.0-42449102270&partnerid=40&md5=ae704eb25ba6e4bbaa98a59b09e43f09 [ * * ( ) ] http://www.scopus.com/inward/record.url?eid=2-s2.0-33847220720&partnerid=40&md5=104123affd3b345af8c2d2bdef599b73 [ * * ( ) ] link:\doibase 10.1007/s00285 - 008 - 0178-y [ * * , ( ) ] link:\doibase 10.1103/physrevlett.110.250601 [ * * , ( ) ] http://www.scopus.com/inward/record.url?eid=2-s2.0-84877749614&partnerid=40&md5=2fea8984888ce36cf51e70c2f0a9509a [ * * ( ) ] link:\doibase 10.1371/journal.pcbi.0020117 [ * * , ( ) ] @noop * * , ( ) @noop `` , '' ( ) @noop `` '' link:\doibase 10.1002/9780470316658 [ _ _ ] , wiley series in probability and mathematical statistics : probability and mathematical statistics ( , , ) pp . , link:\doibase 10.1103/physrevlett.90.020601 [ * * , ( ) ] link:\doibase 10.1103/physrevlett.87.250602 [ * * , ( ) ] \doibase http://dx.doi.org/10.1016/j.cam.2006.02.063 [ * * , ( ) ] , link:\doibase 10.1103/physreve.80.021909 [ * * , ( ) ] link:\doibase 10.1103/physreve.84.021128 [ * * , ( ) ] @noop ( ) link:\doibase 10.1002/9780470316962 [ _ _ ] , ed . , wiley series in probability and statistics : probability and statistics ( , , ) pp . , | delay is an important and ubiquitous aspect of many biochemical processes . for example
, delay plays a central role in the dynamics of genetic regulatory networks as it stems from the sequential assembly of first mrna and then protein .
genetic regulatory networks are therefore frequently modeled as stochastic birth - death processes with delay . here
we examine the relationship between delay birth - death processes and their appropriate approximating delay chemical langevin equations .
we prove that the distance between these two descriptions , as measured by expectations of functionals of the processes , converges to zero with increasing system size .
further , we prove that the delay birth - death process converges to the thermodynamic limit as system size tends to infinity .
our results hold for both fixed delay and distributed delay .
simulations demonstrate that the delay chemical langevin approximation is accurate even at moderate system sizes .
it captures dynamical features such as the spatial and temporal distributions of transition pathways in metastable systems , oscillatory behavior in negative feedback circuits , and cross - correlations between nodes in a network .
overall , these results provide a foundation for using delay stochastic differential equations to approximate the dynamics of birth - death processes with delay . |
from the bethe - weizsacker mass formula , it is well understood that the symmetry energy from bulk matter is the difference between the energy of pure neutron matter and pure symmetric matter .
mathematically , it can we written as @xmath0 where @xmath1 and @xmath2 .
@xmath3 , @xmath4 , and @xmath5 are the neutron , proton , and nuclear matter densities , respectively .
the symmetry energy has great importance in the dense matter existing in the neutron stars , but only indirect information can be extracted from astrophysical observations @xcite .
it is also important in the quark gluon plasma ( qgp ) and hadron gas ( hg ) phase @xcite .
the qgp and hg phases existed in the early stage of the evolution of universe ( about 15 billion years ago ) and are inaccessible nowadays .
it is difficult to recreate these conditions , although numerous experiments are occurring at the relativistic heavy ion collider ( rhic ) and the large hadron collider ( lhc ) @xcite .
heavy - ion reactions , during which matter goes through compression and expansion , are considered to be the true testing ground for the hot and dense matter phases .
the nuclear equation of state ( neos ) and the density dependence of the symmetry energy can be probed by some observables in intermediate - energy heavy - ion collisions ( hics ) .
the softness of the neos has been well described in the literature in the last couple of decades @xcite .
however , the density dependence of the symmetry energy , from the coulomb barrier to the deconfinement of nuclear matter , is a hot topic in the present era . at sub - saturation densities , the density dependence of the symmetry energy
is studied by observables such as the neutron - to - proton ratio , isotopic and isobaric scaling , isospin diffusion , isospin fractionation and/or distillation , and isospin migration .
recently , the msu group @xcite claimed the softness of the symmetry energy at sub - saturation densities by using the double neutron - to - proton ratio and isospin diffusion from two isotopic systems , @xmath6sn + @xmath6sn and @xmath7sn + @xmath7sn at e = 50 mev / nucleon . in another study ,
again soft symmetry energy was claimed by using the isospin diffusion for the same set of reactions , but at e = 35 mev / nucleon @xcite . in a recent study ,
soft symmetry energy is also favored for the same set of reactions at e = 50 mev / nucleon by using the neutron - to - proton ratio @xcite . in all the studies
, the problem of sub - saturation density dependence of the symmetry energy seems to be addressed to some extent ; however , the uncertainties are still large enough to justify the large amount of work that is under way in many laboratories all over the world . in contrast
, the present status of supra - saturation density dependence of the symmetry energy is quite uncertain and interesting .
the high - density behavior of the symmetry energy in the literature is studied by using two important parameters : one is the yield ratio parameter and second is the flow parameter . the yield ratio parameter has been studied in term of single and double ratios of neutrons to protons @xcite , single and double ratios of @xmath8 @xcite , the @xmath9 ratio @xcite , the @xmath10 ratio @xcite , and isospin fractionation @xcite , while , the flow parameter has been studied in terms of relative and differential flows ( single and double ratios ) of neutrons to protons or @xmath11h to @xmath11he @xcite , and in terms of the ratio @xcite or difference @xcite of neutron - to - proton elliptic flow . before using the @xmath11h and @xmath11he particle yield and flow ratios for the density dependence of the symmetry energy at high incident energies
, one must check the production of these particles in the supra - saturation density region , which is obtained during the highly compressed stage only .
however , the production of neutrons and protons occurs in large amounts and can explain the high density dependence of symmetry energy with great accuracy .
favorable results with neutron and proton elliptic flow at e = 400 mev / nucleon were also observed in 2011 . in one of the studies ,
the softness of the symmetry energy with @xmath12 is predicted by comparing the fopi collaboration data with the neutron - to - proton elliptic flow ratio @xcite . in the same year , cozma _ et al .
_ @xcite predicted the softness of symmetry energy with @xmath13 = 2 by comparing the fopi data with the neutron - to - proton elliptic flow difference .
even then uncertainty lies in the results , in terms of determination of symmetry energy : in the first study , symmetry energy is momentum independent , while in later one it is from momentum - dependent interactions .
moreover , the studies were limited to only 400 mev / nucleon .
let us examine some interesting features from the ratio parameters at supra - saturation densities .
all the ratio parameters show sensitivity to the symmetry energy . in the literature
, it is also claimed that @xmath14 and @xmath9 have more sensitivity than @xmath8 @xcite .
the sensitivities of all the parameters is checked in term of transverse momentum and rapidity distribution dependence @xcite , while pion and kaon ratio studies are extended with the isospin asymmetry of the system and the incident energy @xcite . in recent years
, when the pion ratio has been compared with the fopi data by using the two well known models ibuu04 and imiqmd , in terms of isospin asymmetry and incident energy , the predictions for the symmetry energy are found to be totally opposite .
imiqmd predicts stiff symmetry energy ( @xmath15 ) @xcite , while ibuu04 predicts soft symmetry energy ( x=1 ) @xcite . in the present era ,
the @xmath8 ratio is supposed to be a strong candidate for predicting the high - density behavior of symmetry energy . just as for @xmath16 and @xmath17 ,
neutrons and protons are also produced in large amounts up to 1 gev / nucleon .
even around 400 mev / nucleon , the production of neutrons and protons is greater than the production of pions .
unfortunately , the neutron - to - proton ratio parameter in most studies is restricted only with the transverse momentum and kinetic energy dependencies @xcite . to draw a fruitful conclusion in the near future , first
, it is very important to study the isospin asymmetry and incident energy dependencies of the neutron - to - proton ratio , just as in the recent @xmath8 study , and then compare the sensitivity to the symmetry energy from both ratios , as the pion ratio results were recently compared with the fopi experimental findings .
second , one has to avoid choosing randomly any type of fragment to study the supra - saturation density dependence of symmetry energy .
for this , it is important to check whether or not a particular type of fragment is formed in the region @xmath18 , which is simple when one addresses the sub - saturation density dependence of symmetry energy .
finally , with increasing incident energy , the time evolution of the density has a different trend at two extremes : one at the time of maximum compression and the second at the freeze - out time ( t = 200 fm / c ) .
it also becomes interesting to see the different stiffnesses of the symmetry energy dependence of the neutron - to - proton ratio for incident energies at the time of maximum compression and at freeze - out time . in the concluding remarks of this paper ,
we have tried to address the following goals : * to check the sensitivities of different kind of fragments to the high - density behavior of symmetry energy . * to check the behavior of the neutron - to - proton ratio at the time of maximum compression and at freeze - out time . * to check the isospin asymmetry and incident energy dependences of single and double neutron - to - proton ratios from different neutron - rich systems to the high - density behavior of symmetry energy , and then compare the sensitivity of symmetry energy from the neutron - to - proton ratio with that from the pion ratio .
this study is similar to recent studies using the pion ratio @xcite . for the present study ,
the isospin quantum molecular dynamics ( iqmd ) model is used to generate the phase space of nucleons , which is discussed in sec .
ii . the results are discussed in sec .
iii , and we summarize the results in sec .
in the iqmd model @xcite , nucleons are represented by wave packets , just as in the qmd model of aichelin @xcite .
these wave packets of the target and projectile interact via the full skyrme potential energy , which is represented by @xmath19 and is given as : @xmath20 here @xmath21 is the coulomb energy and @xmath22 originates from the density dependence of the nucleon optical potential , and is given as @xmath23 the first two of the three parameters of eq .
[ equation2 ] ( @xmath24 and @xmath25 ) are determined by demanding that , at normal nuclear matter densities , the binding energy should be equal to 16 mev and the total energy should have a minimum at @xmath26 .
the third parameter @xmath27 is usually treated as a free parameter .
its value is given in term of the compressibility : @xmath28 the different values of compressibility give rise to soft and hard equations of state .
the soft equation of state is employed in the present study with the parameters @xmath29 mev , @xmath30 mev , and @xmath27 = 7/6 , corresponding to an isoscaler compressibility of @xmath31 mev . in the third term @xmath32
is the potential part of the symmetry energy , which is adjusted on the basis of calculations from the microscopic or phenomenological many - body theory , having the form @xmath33 here @xmath34mev , parameterized on the basis of the experimental value of the symmetry energy , is known as the symmetry potential energy coefficient . on the basis of the @xmath35 value , symmetry energy is divided into two types with @xmath36 and @xmath37 , corresponding to soft and stiff symmetry energies , respectively . the total symmetry energy per nucleon employed in the simulation is the sum of the kinetic and potential terms and is given as @xmath38 where @xmath39 mev is known as the symmetry kinetic energy coefficient .
the kinetic symmetry energy originates from the fermi - dirac distribution @xcite .
finally , we get a density and isospin - single particle potential in nuclear matter as follows : @xmath40 here @xmath41 , @xmath42 , and @xmath43 .
the potential also depends on the momentum - dependent interactions , which are optional in the iqmd model .
note that the @xmath27 used in the determination of the equation of state and @xmath35 used in the determination of symmetry energy are different parameters .
the interesting feature of symmetry energy is that its value increases with decreasing @xmath35 at sub - saturation densities , while the opposite is true at supra - saturation densities . in other words ,
soft symmetry energy is more pronounced at sub - saturation densities , while stiff symmetry energy is more pronounced at supra - saturation densities . in the calculations
, we use the isospin - dependent in - medium cross section in the collision term and the pauli blocking effects as in the qmd model @xcite .
the cluster yields are calculated by means of the coalescence model , in which particles with relative momentum smaller than @xmath44 and relative distance smaller than @xmath45 are coalesced into a one cluster .
the value of @xmath45 and @xmath44 for the present work are 3.5 fm and 268 mev / c , respectively .
the neutron - to - proton ratio is among the first observables that was proposed as a possible sensitive probe for symmetry energy prediction at sub - saturation densities @xcite ; however , some studies are also performed using the rapidity distribution and transverse momentum dependencies at supra - saturation densities . in this article , the sensitivities of free nucleons , light charged particles ( lcps , having charge number between 1 and 2 ) , and intermediate mass fragments ( imfs , having charge number between 3 and @xmath46 ) to the high density behavior of symmetry energy are checked , providing the results of incident energy and isospin asymmetry dependencies of single and double neutron - to - proton ratios with the high - density sensitive fragments . to perform the study ,
thousand of events are simulated for the isotopes of sn , namely @xmath6sn + @xmath6sn , @xmath7sn + @xmath7sn and @xmath47sn + @xmath47sn between incident energies of 50 and 600 mev / nucleon at semicentral geometry by using the soft and stiff symmetry energies of @xmath36 and 1.5 , respectively . as discussed earlier
, a soft equation of state with an isospin- dependent nucleon - nucleon ( nn ) cross section of @xmath48 is employed .
the incident energy and isospin asymmetry dependences of single and double neutron - to - proton ratios , just as for the @xmath8 ratio @xcite , are considered as a point of importance in the present study .
the single ratio is just the ratio of neutrons to protons and is represented in the study by @xmath49 , while double ratio is the ratio of the single ratios of any two isotopes of sn . in order to study the systematics of the isospin effects , the single ratio of the isotope with a greater number of neutrons is always mentioned in the numerator when the double ratio is calculated .
mathematically , the double ratio is represented by @xmath50 and is given as @xmath51 ( color online ) time evolution of the average density for the @xmath47sn + @xmath47sn reaction with the soft symmetry energy ( @xmath52 ) at semicentral geometry .
the different lines represent different incident energies ranging from 50 to 600 mev / nucleon.,width=340 ] to predict the high - density behavior of symmetry energy , the very first point is to understand the time evolution of the average density at different incident energies . with increasing incident energy , the density will be expected to be greater than the normal nuclear matter density in the most compressed region .
as we know , the density of the environment surrounding the nucleons of a fragment plays a crucial role in determining the physical process behind its formation . in fig .
[ fig:1 ] , we display the average density @xmath53 reached in a typical reaction as a function of time at different incident energies for @xmath47sn + @xmath47sn by using the soft symmetry energy @xmath36 . the average nucleon density is calculated as @xcite @xmath54\rangle,\end{aligned}\ ] ] with @xmath55 and @xmath56 being the position coordinates of the @xmath57 and @xmath58 nucleons , respectively .
as we have expected , with increasing of incident energy , the density is found to increase in the compression zone .
interestingly , at lower beam energies , the maximum density reached is lower and the reaction time is longer . with increasing incident energy , the life - time of the high - density nuclear matter gets shorter due to instability .
for example , at b = 2 fm the average density reaches a maximum and is close to normal nuclear matter density at @xmath59= 18 and 33 fm / c , respectively , for e = 50 mev / nucleon ; but for the case of e = 600 mev / nucleon , the respective times are 10 and 20 fm / c .
this means that the difference between the two times is almost 15 fm / c at e = 50 mev / nucleon , while it is only 10 fm / c at e = 600 mev / nucleon .
this clearly indicates that the matter shows high - density behavior only for a small time interval , which decreases with increasing incident energy
. since we are interested in the sensitivities of different kinds of fragments and their neutron - to - proton ratios , only those fragments that lie in the high - density region ( @xmath18 ) will be sensitive to the high - density behavior of the symmetry energy .
( color online ) time evolution of free nucleons ( top ) , lcps ( middle ) , and imfs ( bottom ) at semicentral geometry for @xmath47sn + @xmath47sn using the soft symmetry energy ( @xmath60 ) .
the different lines have the same meaning as in fig .
[ fig:1 ] .
the vertical line in each panel represents our time limit before which the system can be in the supra - saturation region ( refer to fig .
1).,width=340 ] to check the sensitivities of different kind of fragments , in fig .
[ fig:2 ] we display time evolution of free nucleons ( top ) , lcps ( middle ) , and imfs ( bottom ) at semicentral geometry for incident energies ranging from 50 to 600 mev / nucleon .
the behavior for all kinds of fragments is consistent with the results in the literature @xcite .
the production of free nucleons increases with incident energy , and lcp production decreases after 400 mev / nucleon . in ref .
@xcite , lcp production is correlated with the nuclear stopping and is also found to have a maximum at 400 mev / nucleon .
imf production is found to decrease after 100 mev / nucleon .
this is due to the different origin of the production of imfs as compared to free and lcps . for more details about the incident energy dependence of imfs , see ref .
@xcite .
our main task is to check the sensitivities of the fragments in the high - density region . for this
, we apply the limit that at least one particle must be produced before the time 20 fm / c , because , in an average , after that time the density becomes lower than normal nuclear matter density for all the incident energies under consideration .
the free nucleons are highly sensitive at all the energies .
this is not true for lcps and imfs .
lcps are produced in this region only after the incident energy reaches 200 mev / nucleon .
in contrast , no imfs are produced in the supra - saturation density region .
this means that imfs are not so sensitive to the high - density dependence of symmetry energy ; however , they can be used at sub - saturation and saturation densities @xcite .
interestingly , the single neutron - to - proton ratio from imfs is found to change with the incident energy ( not shown here ) , but this is mainly due to coulomb interactions .
here we conclude that the neutron - to - proton ratio from free nucleons as well as lcps can act as a probe of the high - density behavior of the symmetry energy .
one more interesting observation is obtained from fig .
[ fig:1 ] . with increasing incident energy , the time evolution of the density is exactly opposite during the compression and expansion stages .
that is , in the expansion stage the average density is found to decrease with increasing incident energy , which was earlier increasing in the compressed zone .
now , we have two aspects of the basis of the time evolution of density : one is the compressed - zone time and second is the freeze - out time .
interestingly , if the density behavior is opposite at the two times , then it would supposedly affect the magnitude of the symmetry energy as well as its effect on the nuclear matter during the whole time evolution .
( color online ) excitation function of @xmath61 , which is proportional to @xmath62 , at semicentral geometry .
the left panel is at the time of maximum compression , while the right one is at the freeze - out time .
solid and open circles with a dashed line represent the contributions of the soft ( @xmath35=0.5 ) and stiff ( @xmath35=1.5 ) symmetry energies , respectively , for the system @xmath47sn + @xmath47sn.,width=340 ] to see the virtual change in the symmetry energy due to the change in the density , we display in fig .
[ fig:3 ] the incident energy dependence of @xmath61 , which is proportional to the symmetry energy , at the time of the maximum compression ( left panel ) and at the freeze - out time ( right panel ) . at the time of maximum compression ,
the symmetry energy rises with the increasing incident energy ( increase in density ) .
as the density is more than the normal nuclear matter density in this region , the stiff symmetry energy is stronger than the soft one . with increasing incident energy ( increase in density )
, the stiff symmetry energy is changing drastically , while , the soft symmetry energy shows little change .
this exactly coincided with the ideal picture of density dependence of the symmetry energy . on the other hand ,
if we look at the energy dependence of @xmath61 at @xmath59 = 200 fm / c , the situation is totally different .
the symmetry energy is found to decrease with increasing incident energy ( decrease in density ) .
now the density is lower than the normal nuclear matter density , so the magnitude of the soft symmetry energy is greater than that of the stiff symmetry energy . in other words ,
the supra - saturation ( sub - saturation ) density region is more neutron rich with @xmath63 ( @xmath60 ) .
the effect from the sub- and supra - saturation density behaviors of symmetry energy will compete and contribute in the final observables . due to the different behavior of density at different times ,
it is important to observe the isospin effects at the time of maximum compression and at the freeze - out time to understand the high - density behavior of symmetry energy .
for this purpose , in the coming sections , the incident energy and isospin asymmetry dependencies of the single and double neutron - to - proton ratios from free nucleons and lcps are analyzed .
( color online ) excitation function of the single neutron - to - proton ratio for free nucleons ( top panel ) and lcps ( bottom panel ) at semicentral geometry .
the left and right panels are the same expect they are at the time of maximum compression and at the freeze - out time , respectively .
solid and open circles represent the soft and stiff symmetry energies , respectively .
solid and dashed lines corresponds to the systems of @xmath6sn + @xmath6sn and @xmath7sn + @xmath7sn , respectively .
, width=340 ] in order to address the sensitivity of symmetry energy at the time of maximum compression and at freeze - out time , we display the incident energy and isospin asymmetry dependencies of the single neutron - to - proton ratio at different times in figs .
[ fig:4 ] , [ fig:5 ] and [ fig:6 ] . the left and right panels are at the time of maximum compression and freeze - out time , respectively . in fig .
[ fig:4 ] , the ratios from free nucleons and lcps are displayed in the top and bottom panels .
the many interesting facts are revealed in the figure .
the incident energy dependencies of the ratios are found to be highly sensitive to symmetry energy for the two different times . as we know , the relative strength of symmetry energy is opposite at sub- and supra - saturation densities with @xmath60 and @xmath63 . in the range of 50 - 150 mev / nucleon ,
only the low density part up to about 1.1@xmath26 contributes .
therefore , in the low - energy region , for free nucleons , we can see the high ratio with the soft symmetry energy at both the times under consideration . at and above 200 mev / nucleon , a broad range of densities up to 1.8@xmath26
is involved .
of course , at about 200 mev / nucleon and above , for the behavior of the high - density symmetry energy , there is a combined effect for particles going through both low - and high - density region . at higher energies ,
a higher @xmath64 ratio is observed with the stiff symmetry energy for the neutron rich system
@xmath47sn + @xmath47sn at the time of maximum compression , which is true with the soft symmetry energy at the freeze - out time .
the result is similar for free nucleons and lcps .
however , lcps are not as sensitive and the ratio is even less than the ratio of the system .
this is due to the excess number of protons involved in the production of lcps compared to free nucleons .
these protons will lower the ratio for lcps .
it is clearly visible that the ratio at both times is almost the same with the stiff symmetry energy , but changes drastically with the soft symmetry energy .
this is due to the fact that , at the time of maximum compression , the density is in the supra - saturation region and the stiff symmetry energy is much higher ( see fig.[fig:3 ] ) than the soft symmetry energy .
therefore , the stiff symmetry energy is able to separate most of the neutrons near the time of maximum compression and then accelerate the neutrons toward higher kinetic energy at later times .
however , the soft symmetry energy is not so high , and the separation of the neutrons takes place for a longer time .
after 50 - 60 fm / c ( see fig .
[ fig:1 ] ) , the density drops to the sub - saturation density region and now the soft symmetry energy has a quite high magnitude ( see fig .
[ fig:3 ] ) compared to the stiff one .
the soft symmetry energy in this region is still separating the neutrons as well as accelerating them high kinetic energy .
that is why the ratio with the soft symmetry energy drastically changes when one goes from compression to freeze - out time , but remains almost constant with the stiff symmetry energy .
mainly , the neutron - to - proton ratio is found to decrease with the incident energy for free nucleons as well as for lcps , just like the @xmath8 ratio .
the decrease in the ratio may be due to two reasons : * one reason may be the role of coulomb interactions with incident energy . with increasing incident energy , chances of break - up of initial correlations among the nucleons becomes stronger , and the production of free nucleons including neutrons and protons will increase .
however , at very low incident energy , the production of neutrons is more due to the symmetry energy because of its repulsive ( attractive ) nature for neutrons ( protons ) . in short ,
due to coulomb interactions , a shift of protons takes place from low to high incident energies .
the effect of the coulomb interactions can be checked by taking the double ratio , which is discussed in fig .
[ fig:7 ] . *
the contribution of pions from secondary - chance nucleon - nucleon collisions might increase with the beam energy .
if a first - chance nucleon - nucleon collision converts a neutron to a proton by producing a @xmath16 , then subsequent collisions of the energetic protons can convert them back to neutrons by producing a @xmath17 .
therefore , at sufficiently high energy , the neutrons , which are produced due to symmetry energy , are changing into the protons and further producing @xmath65 s , which will lead to a decease in the neutron - to - proton ratio .
this can be confirmed by using the double ratio concept .
if the double ratio is still deceasing with incident energy , then it means that , in addition to the coulomb interactions , the phenomenon of secondary nucleon - nucleon collisions is also very important .
one more point of interest is that the difference between the soft and stiff symmetry energies at freeze - out time is found to decrease with incident energy for free nucleons , while it increases for lcps . of the above two reasons , the first one is applicable for free nucleons as well as for lcps .
the second one is applicable only for free nucleons , as the energy in this study is up to 600 mev / nucleon , which is quite sufficient to produce pions . to see the effect of the high - density behavior of symmetry energy on the isospin asymmetry dependence
, we display the ratio from free nucleons and lcps in figs .
[ fig:5 ] and [ fig:6 ] at only high energies ( 200 , 400 , and 600 mev / nucleon ) . due to the instability of the highly compressed zone
, we are not able to differentiate between the results of symmetry energy obtained at the time of maximum compression ; however , we had earlier obtained some important conclusions from fig .
[ fig:4 ] , where incident energy dependence was discussed . the results from figs .
[ fig:5 ] and [ fig:6 ] at the freeze - out time reveal many important points .
the isospin asymmetry dependence of the ratio from free nucleons is highly sensitive to the symmetry energy compared to lcps , i.e. , the ratio from free nucleons is found to be sharply increasing with the isospin asymmetry of the system compared to lcps .
this is due to the fact that isospin effects on the ratio from free nucleons are strongly affected by the symmetry energy and weakly affected by coulomb interactions , while the opposite is true for the ratio from lcps .
as discussed earlier , the ratio is found to decease with the incident energy , which is also true here for the isospin asymmetry dependence .
the difference between the soft and stiff symmetry energy results comes from the behavior of free nucleon emission with the isospin asymmetry of the system , i.e. , the greater the isospin asymmetry of the system , the greater the contribution of neutrons in the ratio due to the symmetry energy . the soft symmetry energy is stronger at the freeze - out time , which will lead to an increase in the ratio more sharply than the stiff symmetry energy .
this effect is again weakly observed in the ratio from lcp s .
isospin asymmetry dependence of the single neutron - to - proton ratio for free nucleons at different incident energies .
the left panel is at the time of maximum compression , while the right panel is at the freeze - out time .
solid and open circles represent the soft and stiff symmetry energies , respectively .
, width=340 ] same as in fig .
[ fig:5 ] but for the lcps.,width=340 ] ( color online ) excitation function of the double neutron - to - proton ratio from different isotopes of sn for free nucleons ( top ) and lcps ( bottom ) .
the vertical line in the bottom panel represent the energy limit above which @xmath50 of lcps becomes more or less insensitive .
solid and open circles represent the soft and stiff symmetry energies , respectively .
the solid , dashed , and dot - dashed line corresponds to double ratios from @xmath47sn + @xmath47sn to @xmath7sn + @xmath7sn , @xmath7sn + @xmath7sn to @xmath6sn + @xmath6sn , and @xmath47sn + @xmath47sn to @xmath6sn + @xmath6sn , respectively .
, width=340 ] isospin asymmetry dependence of the double neutron - to - proton ratio from free nucleons at different incident energies .
the different symbols have the same meaning as in fig . [ fig:5 ] .
, width=340 ] incident energy dependence of the power - law exponent @xmath66 from fig .
[ fig:8 ] .
the symbols and lines are the same as in figs . 5 and 8.,width=340 ] in order to cancel the coulomb effects and to see the effect of symmetry energy , we show in fig .
[ fig:7 ] , the incident energy dependence of the double ratio from different isotopes of sn with different combinations , namely , @xmath47sn + @xmath47sn and @xmath7sn + @xmath7sn , @xmath7sn + @xmath7sn and @xmath6sn + @xmath6sn , @xmath47sn + @xmath47sn and @xmath6sn + @xmath6sn , having differences of 8 , 12 , and 20 neutrons and the same number of protons .
the upper and lower panels are for free nucleons and lcps .
the double ratio is found to increase with the difference of the number of neutrons in the different combinations .
it is similar to the results obtained at sub - saturation densities by different models @xcite .
the main point to be discussed here is that the double ratio is found to decrease with increasing incident energy . as we have discussed in fig .
[ fig:4 ] , there may be two reasons for the decrease of the single ratio with increasing incident energy .
one was the coulomb effect , which is canceled out here .
the second was the pion effect , which is still active in the double ratio and becomes more and more dominant with increasing incident energy . due to that effect
, the double ratio is found to decrease with the incident energy .
it indicates that the pion production effect is very important at high incident energy and is equally useful for understanding the high - density behavior of the incident energy @xcite .
in contrast , this effect is valid only for the double ratio from the free nucleons and not from the lcps . the double ratio from lcps is found to be constant above 200 mev / nucleon .
this indicates that the effect of the symmetry energy for the ratio from lcps can be analyzed only near sub - saturation densities close to 1.1@xmath26 .
the decrease in the single ratio for the lcps was only due to the coulomb interactions at higher incident energies , which is canceled out by taking the double ratio ; the double ratio from the symmetry energy becomes independent of the incident energy after 200 mev / nucleon .
this type of dependence for the single @xmath8 ratio can be observed above 1 gev / nucleon @xcite .
the behavior of symmetry energy for the double ratio is exactly the same as that for the single ratio .
this indicates that lcp production is also not a sensitive probe for investigating the high - density behavior of the symmetry energy .
the only possible probe from the fragments is the double ratio of neutrons to protons from free nucleons .
another possible probe is the @xmath8 ratio , which recently was compared with the experimental data of the fopi by the ibuu04 and imiqmd calculations @xcite . in order to strengthen our conclusion , in fig .
[ fig:8 ] , we display the isospin asymmetry dependence of the double ratio from free nucleons at different incident energies .
all the curves are fitted with a power law of the form @xmath67 , where @xmath68 is the double ratio from free nucleons and @xmath13 is the double ratio of the systems .
the power - law exponent @xmath66 is found to vary drastically with the symmetry energy , which is to be discussed later in fig .
[ fig:9 ] . after canceling the coulomb effects
, the trend for the double ratio is the same as that of the single ratio in fig .
[ fig:5 ] .
it reflects the fact that the isospin effects for free nucleons is stronger for more neutron - rich systems and is mainly due to the symmetry energy .
however , the decrease in the isospin effect with the increase of incident energy is due to the production of pions at sufficiently high energy .
the difference in the double ratio obtained with the soft and stiff symmetry energies here is also found to increase from the neutron - poor to the neutron - rich system , just like the single neutron - to - proton ratio in fig .
[ fig:5 ] as well as the single pion ratio in the literature @xcite . to see the clear systematics of the incident energy toward the symmetry energy , we plot the incident energy dependence of the power exponent @xmath66 in fig .
[ fig:9 ] , which is extracted from the curves of fig .
[ fig:8 ] . with increasing incident energy ,
the sensitivity of the symmetry energy goes on decreasing toward the double ratio ; however , the soft symmetry energy is more sensitive in comparison with the stiff one . in brief ,
when one goes from the sub - saturation to the supra - saturation density region , the soft symmetry still has a crucial role to play compared to the stiff one .
this is due to the density ( fig .
[ fig:1 ] ) , which undergoes a sudden change between the supra- and sub - saturation density regions with time at higher incident energies .
finally , from this study , we confirm that the high - density behavior of symmetry energy can be studied by using the single and double ratios of neutrons to protons from free nucleons . in comparison ,
the double ratio is more accurate for this purpose , due to its greater sensitivity to the soft symmetry energy .
meanwhile , the lighter and heavier fragments ratio can be considered good candidates at sub - saturation densities , and also have been used in the literature many times by different groups @xcite .
in order to investigate the high - density behavior of the symmetry energy , isospin asymmetry and beam energy dependencies of neutron - to - proton ratios ( single and double ) from different kinds of fragments are studied by using the iqmd model . the single neutron - to - proton ratio from free nucleons and lcps
is found to decrease ( increase ) with incident energy ( with the isospin asymmetry of the system ) .
stronger isospin effects are observed with the soft symmetry energy .
similar results with the @xmath8 ratio are also observed by li _
et al . _ and feng _ et al . _ , but with opposite behavior for symmetry energy .
the double neutron - to - proton ratio from free nucleons is highly sensitive to the symmetry energy , incident energy , and isospin asymmetry of the system .
however , the sensitivity of the neutron - to - proton double ratio from lcps to the nuclear symmetry energy is almost beam - energy independent above 200 mev / nucleon .
the same trend is observed for the single @xmath8 ratio above 1 gev / nucleon .
the sensitivity of the soft symmetry energy to the ratio parameter is strongly affected by the choice of times , which is not true for the stiff symmetry energy . in simple words , just like the @xmath8 ratio , the neutron - to - proton double ratio from free nucleons can act as a useful probe to constrain the high - density behavior of symmetry energy .
experiments are planned at msu , gsi , riken , and frib to determine the high - density behavior of symmetry energy by using the neutron - to - proton ratio .
this work is supported in part by the chinese academy of sciences support program for young international scientists under grant no .
2010y2jb02 , the national science foundation of china under contract no . 11035009 , and no . 10979074 , by the the knowledge innovation project of the chinese academy of sciences under grant no .
kjcx2-ew - n01 , and by the 973-program under contract no .
2007cb815004 .
j. m. lattimer and m. prakash , science * 304 * , 536 ( 2004 ) ; a. w. steiner _
et al . _ ,
phys . rep . *
411 * , 325 ( 2005 ) ; b. a. li , l. w. chen , m. gearheart , j. hooker , c. m. ko , p. g. krastev , w. k. lin , w. g. newton , d. h. wen , c. xu , and j. xu , j. phys . conf . ser . * 312 * , 042006 ( 2011 ) ; d. t. loan _
rev c * 83 * , 065809 ( 2011 ) .
m. d. toro , b. liu , v. greco , v. baran , m. colonna , and s. plumari , phys .
c * 83 * , 014911 ( 2011 ) .
j. l. nagle , i. g. bearden , and w. a. zajc , new j. phys .
* 13 * , 075004 ( 2011 ) .
j. aichelin , phys . rep . * 202 * , 233 ( 1991 ) ; e. lehmann , r. k.puri , a.faessler , t. maruyama , g. -q .
li , n. ohtsuka , s. w. huang , d. t. khoa _ et al .
. nucl .
* 30 * , 219 ( 1993 ) ; e. lehmann , r. k. puri , a. faessler , g. batko , s. w. huang , phys .
c * 51 * , 2113 ( 1995 ) ; r. k. puri , and j. aichelin , j. comput .
* 162 * , 245 ( 2000 ) ; d. t. khoa , n. ohtsuka , m. a. matin , and r. k. puri , nucl .
* a548 * , 102 ( 1992 ) .
p. danielewicz , r. lacey , and w. g. lynch , science * 298 * , 1592 ( 2002 ) ; c. hartnack , h. oeschler , and j. aichelin , phys . rev . lett . * 96 * , 012302 ( 2006 ) ; z. q. feng , phys .
c * 83 * , 067604 ( 2011 ) .
b. a. li , c.m .
ko , and z.z .
ren , phys .
lett . * 78 * , 1644 ( 1997 ) ; h. s. xu _
et al . _ ,
lett . * 85 * , 716 ( 2000 ) ; m. b. tsang _
et al . _ , phys . rev .
lett . * 86 * , 5023 ( 2001 ) ; m. b. tsang _
et al . _ ,
lett . * 92 * , 062701 ( 2004 ) ; d. v. shetty , s. j. yennello , a. s. botvina , g. a. souliotis , m. jandel , e. bell , a. keksis , s. soisson , b. stein , and j. iglio , phys .
c * 70 * , 011601 ( 2004 ) ; v. baran , m. colonna , m. d. toro , m. z. pfabe , and h. h. wolter , phys.rev .
c * 72 * , 064620 ( 2005 ) ; m. a. famiano , t. liu , w. g. lynch , a. m. rogers , m. b. tsang , m. s. wallace , r. j. charity , s. komarov , d. g. sarantites , and l. g. sobotka , phys .
lett . * 97 * , 052701 ( 2006 ) ; w. d. tian , y. g. ma , x. z. cai , d. q. fang , w. guo , w. q. shen , k. wang , h. w. wang , and m. veselsky , phys .
c * 76 * , 024607 ( 2007 ) ; m. colonna , v. baran , m. d. toro , and h. h. wolter , phys .
c * 78 * , 064618 ( 2008 ) ; m. d. toro _ et al .
e * 17 * , 1799 ( 2008 ) ; i. lombardo _
et al . _ ,
phys . rev .
c * 82 * , 014608 ( 2010 ) .
y. zhang , p. danielewicz , m. famiano , z. li , w.g .
lynch , and m.b .
tsang , phys .
b * 664 * , 145 ( 2008 ) ; m. b. tsang , y. zhang , p. danielewicz , m. famiano , z. li , w. g. lynch , and a. w. steiner , phys .
lett . * 102 * , 122701 ( 2009 ) ; m. b. tsang _
. phys . * 66 * , 400 ( 2011 ) .
z. y. sun _ et al .
c * 82 * , 051603 ( r ) ( 2010 ) .
s. kumar , y. g. ma , g. q. zhang , and c. l. zhou , phys .
c * 84 * , 044620 ( 2011 ) .
b. a. li , phys .
* 88 * , 192701 ( 2002 ) ; nucl .
a * 708 * , 365 ( 2002 ) .
b. a. li , c. b. das , s. d. gupta , and c. gale , phys .
c * 69 * , 011603 ( r ) ( 2004 ) ; b. a. li , _ ibid . _ * 69 * , 034614 ( 2004 ) ; _ ibid . _
* 71 * , 044604 ( 2005 ) ; b. a. li , l. w. chen , g. c. yong , and w. zuo , phys .
b * 634 * , 378 ( 2006 ) ; q. li , z. li , s. soff , m. bleicher , and h. stcker , phys .
c * 72 * , 034613 ( 2005 ) .
b. a. li , l. w. chen , h. r. ma , j. xu , and g. c. yong , phys .
c * 76 * , 051601 ( r ) ( 2007 ) .
b. a. li , phys .
c * 67 * , 017601 ( 2003 ) .
q. li , z. li , e. zhao , and r. k. gupta , phys .
c * 71 * , 054907 ( 2005 ) .
h. h. wolter , v. prassa , g. lalazissis , t. gaitanos , g. ferini , m. d. toro , and v. greco , prog .
62 * , 402 ( 2009 ) .
z. g. xiao , b. a. li , l. w. chen , g. c. yong , and m. zhang , phys .
lett . * 102 * , 062502 ( 2009 ) ; m. zhang , z. g. xiao , g. c. yong , b. a. li , l. w. chen , and s. j. zhu , phys .
c * 80 * , 034616 ( 2009 ) . z.
q. feng and g. m. jin , phys .
b * 683 * , 140 ( 2010 ) .
m. zhang , z. g. xiao , and s. j. zhu , chin .
c * 34 * , 1100 ( 2010 ) ; phys .
c * 82 * , 044602 ( 2010 ) ; y. gao , l. zhang , h. f. zhang , x. m. chen , and g. c. yong , phys .
c * 83 * , 047602 ( 2011 ) .
g. c. yong , b. a. li , and l. w. chen , phys .
c * 74 * , 064617 ( 2006 ) ; g. c. yong , b. a. li , l. w. chen , and x. c. zhang , _ ibid .
_ * 80 * , 044608 ( 2009 ) ; g. c. yong , b. a. li , and l. w. chen , int .
e * 19 * , 1647 ( 2011 ) ; s. gautam , a. d. sood , r. k. puri , and j. aichelin , phys .
c * 83 * , 034606 ( 2011 ) .
p. russotto _
et al . _ ,
b * 697 * , 471 ( 2011 ) .
m. d. cozma , phys . lett .
b * 700 * , 139 ( 2011 ) . c. hartnack , r. k. puri , j. aichelin , j. konopka , s. a. bass , h. stoecker , and w. greiner , eur .
j. a * 1 * , 151 ( 1998 ) ; c. hartnack , h. oeschler , y. leifels , e. l. bratkovskaya and j. aichelin , phys .
* 510 * , 119 ( 2012 )
. s. kumar , s. kumar , r. k. puri , phys .
rev c * 81 * , 014601 ( 2010 ) ; _ ibid .
_ * 81 * , 014611 ( 2010 ) ; s. kumar and s. kumar , chin
. phys . lett .
* 27 * , 062504 ( 2010 ) ; s. gautam and a. d. sood , phys
c * 82 * , 014604 ( 2010 ) ; s. kumar , rajni , s. kumar , phys .
c * 82 * , 024610 ( 2010 ) ; s. kumar , v. kaur , and s. kumar , cent .
* 9 * , 986 ( 2011 ) .
q. f. li , ph.d .
thesis , ciae , beijing , china , 2002 .
s. kumar , s. kumar , and r. k. puri , phys .
c * 78 * , 064602 ( 2008 ) ; s. kumar and s. kumar , pramana j. of phys .
* 74 * , 731 ( 2010 ) . | the symmetry energy at sub- and supra - saturation densities has great importance for understanding the exact nature of asymmetric nuclear matter as well as neutron stars , but it is poorly known , especially at supra - saturation densities .
we will demonstrate here whether or not the neutron - to - proton ratios from different kinds of fragments can determine the supra - saturation behavior of the symmetry energy .
for this purpose , a series of sn isotopes were simulated at different incident energies using the isospin quantum molecular dynamics ( iqmd ) model with either a soft or a stiff symmetry energy .
it is found that the single neutron - to - proton ratio from free nucleons as well as light charged particles ( lcps ) is sensitive to the symmetry energy , incident energy , and isospin asymmetry of the system .
however , with the double neutron - to - proton ratio , this is true only for the free nucleons .
it is possible to study the high - density behavior of symmetry energy by using the neutron - to - proton ratio from free nucleons . |
core - collapse supernovae ( sne ) mark the explosive deaths of massive stars .
several independent lines of evidence including explosion modeling @xcite , progenitor star photometry @xcite , rate statistics @xcite , and theory @xcite combine to suggest a lower main sequence initial mass ( @xmath2 ) limit for achieving core collapse of @xmath3 .
red supergiant progenitor stars in this mass range are known to produce type iip ( hydrogen rich ) sn explosions , the most common form of core - collapse sn .
the upper mass limit for sne iip progenitors is more uncertain , with stars of @xmath4 realizing significant mass loss depending on their mass , metallicity , rotation rate , binarity , and other properties ; and even more massive stars ending their lives through more exotic explosion mechanisms .
these mass limits for cc - sn progenitor stars have profound implications throughout stellar and galactic astrophysics and cosmology , including as an input to and constraint on models of stellar evolution for massive stars @xcite , chemical evolution @xcite , supernova feedback in the interstellar medium and galaxy formation @xcite , and astrobiological planetary sterilization rates @xcite .
the electromagnetic signatures of these core - collapse explosions are diverse , depending sensitively on the properties of both the core and the outer envelope of the progenitor star at the time of explosion .
supernovae with hydrogen features detected in their optical spectra are referred to as type ii sne , with a variety of subtypes defined by more specific spectroscopic and/or photometric criteria ( see e.g. * ? ? ?
* ; * ? ? ?
the most common subclass , type iip , are typified by broad ( @xmath5 ) hydrogen balmer p - cygni spectroscopic features , fast rise times of a few days and optical light curves dominated by a long lived , @xmath6 day `` plateau '' phase of roughly constant luminosity .
the plateau phase is understood to arise from hydrogen recombination in the ejecta , with cooling temperature balancing the expansion of the blastwave to essentially equilibrate the @xmath7-band luminosity ( see e.g. * ? ? ?
the type iil sub - class is historically designated based on spectroscopic properties similar to sne iip , but faster , `` linearly '' declining optical light curves rather than a long lived plateau .
type iib supernovae are classified spectroscopically based on the disappearance of h features and the prominence of he absorptions .
type iib light curves feature slow rise times and rapid decline rates ( in each case , a few weeks ) typical of type i ( h deficient ) sne .
the most extreme subclass , type iin , are identified by intermediate width ( @xmath8 ) h emission features reflecting interaction of supernova ejecta with circumstellar material , and contributions from this interaction can power these explosions to reach extreme luminosities at peak . the optical evolution of type iip sne has been explored in light curve studies by a number of authors , including @xcite .
the relationship between these observables and the properties of sn progenitor stars has been explored in theoretical parameter studies by @xcite , and others . combining a uniform analysis of a statistical population of type iip supernova light curves with consistent physical models for inferring the properties of their stellar progenitors represents a path forward for characterizing the progenitor population .
here we describe an analysis of a statistical sample of sn iip light curve properties , performed using observations from the panoramic survey telescope & rapid response system 1 survey ( , abbreviated ps1 ) .
this represents the first such population analysis of sn iip light curves based on a homogeneously - collected and multi - band photometric sample from a wide field optical survey . in section [ sec : obs ] we describe the ps1 optical observations and follow - up optical spectroscopy program used to construct the light curve sample .
we have developed a novel bayesian methodology for self - consistently modeling the full population of light curves in the sample and obtaining robust measurements of physically - meaningful light curve parameters ( section [ sec : model ] ) .
we discuss the population wide distributions of these parameters and compare them to previous observational studies ( section [ sec : res ] ) . by comparison to theoretical light curve models ,
we recover estimates of the progenitor properties of the objects in our sample , and discuss the limitations of the available models in section [ sec : prog ] .
we summarize and conclude in section [ sec : conc ] . in a companion paper , @xcite
, we apply the ps1 sn ii data presented here as a test case for a hierarchical bayesian light curve fitting methodology which enables simultaneous modeling of full populations of transient light curves .
we select a type ii sn light curve sample from the transients discovered and monitored by ps1 since the initiation of the survey in 2010 , consisting of 18,953 relevant photometric data points , 5,096 of which are robust transient detections .
ps1 is a high - etendue wide - field imaging system , designed for dedicated survey observations and located on the peak of haleakala on the island of maui in the hawaiian island chain .
routine observations are conducted remotely , from the university of hawaii
institute for astronomy advanced technology research center ( atrc ) in pukalani .
a summary of details of ps1 operations relevant to sn studies is given in @xcite , and we discuss its key features here . a complete description of the ps1 system , both hardware and software , is provided by @xcite . the 1.8 m diameter primary mirror , @xmath9 field of view , and other ps1 optical design elements are described in @xcite ; the array of sixty @xmath10 , 0.258pixel detectors , and other attributes of the ps1 imager is described in @xcite ; and the survey design and execution strategy are described in @xcite . the ps1 medium deep survey ( mds ) consists of 10 pencil beam fields observed with a typical cadence of 3 days in each filter , to a @xmath11 depth of @xmath12 mag in @xmath13 filters , and @xmath14 mag in the @xmath15-filter ( with observations taken near full moon ) .
the ps1 observations are obtained through a set of five broadband filters , which we refer to interchangeably as as @xmath16 , @xmath17 , @xmath18 , @xmath19 , and @xmath20 or simply @xmath0 @xcite .
although the filter system for ps1 has much in common with that used in previous surveys , such as the sloan digital sky survey ( sdss : * ? ?
* ; * ? ? ?
* ) , there are important differences .
the @xmath16 filter extends @xmath21 redward of @xmath22 , and the @xmath19 filter is cut off at @xmath23 .
sdss has no corresponding @xmath20 filter .
further information on the passband shapes is described in @xcite .
photometry is in the `` natural '' ps1 system , @xmath24 , with a single zero - point adjustment @xmath25 made in each band to conform to the ab magnitude scale @xcite .
we assume a systematic uncertainty of 1% for our ps1 observations due to the asymmetric ps1 point spread function and uncertainty in the photometric zero - point calibration @xcite .
see figure [ fig : mlcsamp ] for an illustration of the ps1 photometric sampling and the range in data quality .
the standard reduction , astrometric solution , and stacking of the nightly images is done by the pan - starrs1 ipp system @xcite .
the nightly mds stacks are transferred to the harvard faculty of arts and sciences `` odyssey '' research computing cluster , where they are processed through a frame subtraction analysis using the _ photpipe _ image differencing pipeline developed for the supermacho and essence surveys @xcite .
we begin with a selection of ps1-discovered sne which were classified as type ii through our spectroscopic follow - up campaign : 112 objects in total .
one object in our sample , ps1 - 10ae ( sn 2010aq ) , has previously been reported on in @xcite .
we note that the ps1 spectroscopic follow - up is not complete ; brighter objects , those with longer plateau durations , and those with the highest ratio of sn to underlying galaxy light are most likely to be over - represented in our sample , due to their availability for spectroscopy .
spectra were obtained using the blue channel and hectospec spectrographs of the 6.5 m mmt telescope ( @xcite ) , the low dispersion survey spectrograph ( ldss3 ) and inamori - magellan areal camera and spectrograph ( imacs ; @xcite ) of the 6.5 m magellan telescopes , the gemini multi - object spectrograph of the 8 m gemini telescopes ( gmos ; @xcite ) , and the and the andalucia faint object spectrograph and camera ( alfosc ) at the 2.6 m nordic optical telescope .
objects were classified as type ii by identification of h@xmath26 emission not associated with the host galaxy .
this selection could possibly include the type iip , iil , iib , and type iin subclasses of the sne ii ; we will discuss sub - classification in section [ sec : mod : class ] .
additionally , we obtain host galaxy redshift measurements for each object from these spectra , which we use for @xmath27-correction ( see section [ ssec : kcor ] ) and to estimate distance.cdm cosmology with @xmath28 km s@xmath29 mpc@xmath29 , @xmath30 , and @xmath31 ; @xcite ] figure [ fig : zdist ] shows the host galaxy redshift distribution of the ps1 sn ii sample , which has [ 16,50,84]th percentile values of @xmath32 $ ] .
the sub - sample classified as sne iip ( see section [ sec : mod : class ] ) have distribution percentile values of [ 0.07,0.10,0.16 ] .
this difference reflects a population of luminous , distant sne iin excluded from the sub - sample .
details of our final sn iip sample , as described in section [ sec : mod : class ] , are listed in table [ tab : obspars ] .
[ sec : obs : ps1snii ] we use @xmath27-corrections to account for the difference between the observed and rest frame wavelengths of the light collected through the ps1 filters , as a function of the redshift of each object in our sample .
we define this correction , @xmath33 , where @xmath34 is the observed - frame magnitude , @xmath35 is the emitted - frame absolute magnitude , and @xmath36 is the distance modulus @xcite .
we estimate @xmath27-corrections using the sn iip spectral templates provided by p. nugent ( based on observations of sn 1999em ; @xcite ) .
we interpolate linearly between the spectral templates to obtain @xmath27-corrections for arbitrary epochs .
we note that we do not warp the spectral templates to match the observed color of the sn , due to the computational cost involved in doing so dynamically within our probabilistic light curve fitting methodology ( see appendix [ ap : stan ] ) . by testing the effects of warping the spectral templates on the @xmath27-correction calculation
, we estimate that neglecting this effect will introduce an uncertainty of @xmath37 mag in the @xmath27-correction per @xmath38 mag of @xmath39 color offset , which is much smaller than typical uncertainties in our color estimates .
figure [ fig : kcor ] illustrates the resulting @xmath27-corrections for an illustrative set of epochs .
the @xmath40 @xmath27-correction is linear with @xmath41 because the post - shock breakout cooling phase spectral model is a blackbody ; thereafter the @xmath27-correction evolution becomes more complex , with spectral features evolving and shifting between bandpasses .
the bluer bands have @xmath27-corrections more highly dependent on time , as the onset of line blanketing in the spectral model depresses the flux blueward of @xmath42 .
deviation of the supernova spectral energy distribution from the template spectra will introduce error into our derived @xmath27-correction .
in particular , we consider here the effect of differences in the sn iip continuum shape and line - of - sight extinction , and estimate the magnitude of these effects by simulation on the template spectra .
first , we consider the effect of variation in continuum shape , to first order .
we vary the intrinsic @xmath39 color of the supernova spectrum by @xmath43 mag from the stock template , at fixed redshift ( @xmath44 ) , epoch ( @xmath45 days ) , and with no extinction .
the effect of this variance in spectral shape on the @xmath27-correction over the full @xmath46 mag color range induces a variation in the @xmath27-correction ranging from @xmath38 mag in @xmath47-band to @xmath48 mag in @xmath15-band .
second , we consider the effect of variation in extinction .
we vary the extinction from @xmath49 to @xmath50 mag , at fixed redshift ( @xmath44 ) , epoch ( @xmath45 days ) , and with the color of the stock template .
the effect of this variance in spectral shape on the @xmath27-correction over the full @xmath50 mag extinction range induces a variation in the @xmath27-correction ranging from @xmath38 mag in @xmath15-band to @xmath51 mag in @xmath52-band .
these effects are therefore modest and similar in size , depending on the exact parameters of the photometric observation .
we estimate the typical effect of the joint variation in both continuum shape and extinction by means of monte carlo simulation .
we generate @xmath53 random supernova observations with the following conditionally independent distributions of properties meant to approximately mimic the ps1 dataset : redshift drawn from @xmath54 , color deviation drawn from @xmath55 mag color distribution is meant to provide a conservative accounting of intrinsic spectral deviance between sne iip . for comparison ,
@xcite identify an intrinsic @xmath39 color dispersion of @xmath56 mag .
] , extinction drawn from @xmath57 mag , and epoch drawn from @xmath58 days , where @xmath59 is the normal and @xmath60 is the lognormal distribution .
we then consider the distribution of residuals of the derived @xmath27-corrections from the @xmath27-correction expected for a zero - extinction and template - standard color object .
in the @xmath61-band , this distribution has @xmath62 $ ] percentile values of @xmath63 $ ] mag .
the distribution is similar in other filters .
this suggests that these effects exert a small ( @xmath64 mag ) typical bias and introduces a modest additional uncertainty ( @xmath65 mag ) in the measured supernova photometry .
we therefore neglect these effects in the following work , but note their importance for studies requiring precision photometry . in the most extreme cases , particularly for the most highly reddened supernovae , the simulations suggest that these effects will cause the brightness to be significantly under - estimated ( by @xmath66 mag ) . moreover , the simulated @xmath27-correction residuals seem to have little correlation between filters , suggesting that the effects do not significantly bias color inference .
in order to consistently compare the photometric properties of the sn iips in our sample , we have produced models of the full , multi - band light curve evolution of each sn using a bayesian methodology .
we apply weakly informative priors to regularize the shape of the light curve fits to conform to sn iip phenomenology .
the product of the modeling is a posterior predictive probability distribution for the luminosity of the sn at every phase , and joint posterior probability distributions for associated light curve parameters .
this methodology enables us to model the full pseudo - bolometric light curve evolution of all objects in our sample from explosion through the radioactive decay phase , regardless of variation in photometric coverage and data quality , while fully accounting for statistical uncertainty . throughout
, we interpret these marginal parameter probability distributions in the context of the influence of the weakly informative priors to avoid introducing bias on our inference . to model the light curve , we use a simple , physically - motivated parameterization that captures the essential components of the rise , plateau , and decline phases of the type iip light curve .
our model is constructed so that the most salient features of the sn iip light curve ( the plateau duration , peak magnitude , and the luminosity evolution of the radioactive decay phase ) are represented by directly interpretable parameters .
the light curve model for the luminosity measured from our ps1 forced photometry , @xmath67 , in a given optical filter , @xmath68 , consists of 5 piecewise components ( see figure [ fig : modschem ] ) : a power law fast rise phase , a plateau phase divided into exponential rising and declining components , a transitional phase representing the end of the plateau , and a post - plateau exponential decay phase : @xmath69=\begin{cases } 0 , \\&\hspace{-14em } \text{if } t < t_0 \\ m_1 ~ ( t / t_1)^{\alpha } , \\&\hspace{-14em } \text{if } t_0<t < t_0+t_1
\\ m_1 \exp(\beta_1 ( t - t_1 ) ) , \\&\hspace{-14em } \text{if } t_0+t_1<t < t_0+t_1+t_p \\ m_p \exp(-\beta_2 ( t-(t_p+t_1 ) ) ) , \\&\hspace{-14em } \text{if } t_0+t_1+t_p < t < t_0+t_1+t_p+t_2 \\ m_2 \exp(-\beta_{dn } ( t-(t_2+t_p+t_1 ) ) ) , \\&\hspace{-14em } \text{if } t_0+t_1+t_p+t_2<t\\&\hspace{-10em}<t_0+t_1+t_p+t_2+t_d \\ m_d \exp(-\beta_{dc } ( t-\\\hspace{4em}(t_d+t_2+t_p+t_1 ) ) ) , \\&\hspace{-14em } \text{if }
t_0+t_1+t_p+t_2+t_d < t \\ \end{cases } \notag \\\end{aligned}\ ] ] these parameters have the following definitions and interpretations .
the time parameters are defined where @xmath70 is the modified julian date ( mjd ) epoch of an observation , @xmath71 is the epoch of explosion , @xmath72 is the rest frame duration of the power law rise phase , @xmath73 is the duration of the exponential rise phase ( ending at peak flux ) , @xmath74 is the duration of the falling component of the plateau phase , and @xmath75 is the duration of the transitional phase .
the flux parameters are defined such that @xmath76 is the flux at the transition from the power law to the exponential rise phases , @xmath77 is the peak flux , @xmath78 is the flux at the end of the plateau phase , and @xmath79 is the flux at the transition to the co decay - dominated phase .
the rate parameters are defined such that @xmath26 is the power law rise slope , @xmath80 is the exponential rate constant during the rising phase of the plateau , @xmath81 is the rate constant during the declining phase of the plateau , @xmath82 is the exponential decline rate of the transition phase following the plateau , and @xmath83 is the exponential decay constant corresponding to @xmath84co to @xmath84fe decay .
each parameter is defined independently for each photometric filter , with the exception of @xmath71 . for numerical convenience ,
we define @xmath67 in arbitrary scaled units relative to the absolute magnitude @xmath85 , such that @xmath86 .
note that , in order for the light curve model to be continuous , not all of the parameters may be independent . in particular
, for each filter , @xmath87 furthermore , we note that the post - peak decay rate @xmath81 is directly related to the quantity @xmath88 , the decline in magnitudes of the light curve in the 15 days following peak : @xmath89 our model is similar to the linear segmented light curve fitting approach of e.g. @xcite , but ( as we will discuss in section [ sec : mod : fit ] ) our fitting methodology for the knot properties is fully probabilistic rather than manual .
we prefer this piecewise analytic formulation to the additive components model used by e.g. @xcite .
the piecewise parameters will in principal have weaker covariance , therefore reducing the posterior curvature and increasing the efficiency of markov chain monte carlo ( mcmc ) methods for sampling from the posterior . while this parameterization is designed to capture the phenomenology of type iip sne , it is in practice quite flexible , as we obtain reasonably descriptive fits to the light curves of other type ii sne ( e.g. sne iin and iib ) .
we estimate the posterior distributions of these model parameters using a mcmc method .
we employ the c++ library _ stan _ @xcite , which implements the adaptive hamiltonian monte carlo ( hmc ) _ no - u - turn sampler _ of @xcite . for each multi - band sn light curve , we use _ stan _ to return 1000 samples ( 250 samples each from 4 independent mcmc chains ) from the posterior distribution of the model .. ] in addition to the light curve parameterization outlined in section [ sec : model ] , our _ stan _ model includes certain features representing the data acquisition process . to account for uncertainty in the ps1 background template subtractions ,
we fit for the background level in each filter using an independent set of luminosity parameters , @xmath90 $ ] , and an intrinsic model variance , @xmath91 $ ] .
we pre - compute @xmath27-correction curves for the redshift of each object in our sample ( see section [ ssec : kcor ] ) , and apply them to the model during the likelihood calculation using the phases corresponding to the sampled explosion date at each step in the mcmc chain .
we employ weakly informative priors ( see e.g. @xcite ) to regularize the fitted models to the characteristic sn iip light curve shape .
these prior distributions are constructed to more than encompass the observed variance in parameter values from well - identified light curves in the dataset .
_ model in appendix [ ap : stan ] fully defines the fixed prior distributions we employ , and we discuss their most salient properties here . figure [ fig : pfprior ] illustrates the procedure for constructing the prior distributions which we have defined individually per filter in the model . for @xmath74 , we apply a ( lognormal ) @xmath92 day prior .
together with our prior on @xmath73 ( figure [ fig : pfprior ] ) , this effectively provides a prior on the plateau duration which has @xmath93$]th percentile values of @xmath94 $ ] days in @xmath52-band ( and similar in other bands ) .
this prior is constructed to be weakly informative , having significantly more variance than previously reported sn iip plateau duration distributions @xcite and covering a range of values similar to that predicted from theoretical modeling @xcite .
perhaps the most informative prior we apply is that on the peak - magnitude distribution , which is observationally motivated .
a strong prior is desirable for this parameter because it has the effect of regularizing the light curve shape in cases where photometric observations before peak are not available .
we set this prior to @xmath95 to approximately match the type ii sn luminosity function predicted for an @xmath7-band magnitude - limited survey by @xcite based on observations from the lick observatory supernova search ( loss ; figure [ fig : mpprior ] ) .. here we assume a uni - modal gaussian prior wide enough to encompass both modes of the @xcite luminosity function , and investigate the possibility of multi - modality in the observed properties of the sne ii in section [ sec : res : iil ] . ] in total , including the independent parameters for each light curve segment and filter , the fitted model for each sn has 61 independent parameters ( 12 per filter and @xmath71 ) .
we reproduce the fitted light curve parameters for each object in our sample in table [ tab : lcpars ] . following execution of _ stan _ and posterior sampling , we reconstruct a posterior predictive distribution for each light curve from the model parameter samples @xcite .
this distribution represents the probability for the physical supernova light curve to have the luminosity @xmath96 on a grid of times @xmath97 .
figure [ fig : modmlc ] illustrates a posterior predictive check for the model fit to the object ps1 - 11apd , whose well - sampled observations illustrate both the strengths and weaknesses of the 5-component model .
the rising phase in each band is well - constrained by the tightest pre - explosion limit , in the @xmath15-band .
the two - component ( rising then falling ) plateau phase provides an accurate description of the data in the @xmath52-band , but not in the @xmath47-band , where the evolution of iron line blanketing ( see e.g. @xcite ) causes a behavior not captured by the model ( falling then rising / plateauing ) . as is characteristic of the ps1 data , the @xmath15-band light curve is more poorly sampled and higher in variance than the other bands . in general , the model parameter inferences are therefore more reliable and directly interpretable in the @xmath61 , @xmath52 , and @xmath41 bands . figure [ fig : modex ] illustrates fitting behavior for a well - sampled and a poorly - sampled ps1 light curve .
the top panel shows ps1 - 12bku , a sn with observations from the rise phase through the end of the plateau .
note that the nonlinear behavior of the light curve during the rise , due to shock breakout and post - shock breakout cooling envelope effects ( see e.g. @xcite ) are not captured by the power law rise model , but the explosion epoch itself is fit accurately in the fitted model .
ps1 - 10zu is shown in the bottom panel , an object which apparently exploded during the gap between observing seasons . with observations only available starting a few weeks
before the end of the plateau phase , and no direct constraints on the explosion epoch , the model fit includes a range of possible explosion dates reflecting the prior distributions for the plateau duration and peak magnitude .
we apply a supervised machine learning methodology ( see e.g. @xcite ) to classify the type ii sne in this sample into their appropriate subclasses and , in particular , to identify sne iip for further analysis .
our approach is to first apply `` expert knowledge '' to manually label a subset of clearly - classifiable objects from our sample of 112 sne ii , then we uniformly measure light curve features using the methodology described in section [ ssec : mod ] , then train a support vector machine ( svm ) classifier to a deliberately - selected subset of these features , and finally apply that classifier to assign labels to the remaining objects in the sample . in general , spectral classification of distant ( @xmath98 ) transients
is complicated by spectral evolution ( see e.g. @xcite for a discussion of sne iib ) , exacerbated by the observational cost of obtaining a well sampled spectral series , and the difficulty in robustly identifying spectral features , exacerbated by the low signal - to - noise of the spectroscopy relative to nearby objects . the methodology discussed below is intended to capitalize on photometric information to help overcome these limitations .
our procedure is as follows .
we begin by manually assigning sn iip , iib , iin , and `` ii ? '' ( undetermined ) classifications to each object in the sample by inspection of their light curves and optical spectra .
see e.g. @xcite for a discussion of the photometric and spectroscopic properties of sn ii sub - classes .
we do not establish a label for sne iil separate from sne iip , but discuss that sub - class later in section [ sec : res : iil ] .
we assign sn iin classifications to 23 objects based on the detection of intermediate - width h emission features and/or exceptional luminosity ( @xmath99 mag ) ; sn iib classifications to 5 objects based on the detection of he features ( in addition to broad h features ) in the optical spectrum and/or a sn i - like light curve , with slow rise and fast decline rate ; and sn iip to 67 objects based on the presence of a broad @xmath100 spectral feature and/or a clearly - exhibited plateau - like light curve .
finally , we liberally apply the sn ii ?
classification to 17 objects for which we do not have sufficient data to assign a classification and/or do not neatly fit the preceding criteria .
we apply our bayesian light curve model ( section [ ssec : mod ] ) to uniformly measure features of the light curves of every sn ii in our ps1 sample . to mirror our expert knowledge , we focus on the peak absolute magnitude ( @xmath85 ) , the rise time to peak magnitude ( @xmath73 ) , and the post - peak decay rate ( @xmath81 ) as features capable of distinguishing objects in these subclasses .
we identify the @xmath52 band as the filter in which these features are most discriminating . for self - consistency
, we apply the sn iip theoretical @xmath27-corrections ( section [ ssec : kcor ] ) to every object in the sample .
we then fit a linear svm classifier @xcite to the features of the training set .
the svm consists of a series of hyperplanes that optimally divide the feature space to discriminate the training set based on their class labels .
the results of this fit are displayed in figure [ fig : svm ] . in summary ,
the classifier identifies sne iin as luminous ( primarily @xmath101 mag ) and slowly - evolving ( @xmath102 ) , sne iib as intermediate - luminosity and quickly evolving , and sne iip as slowly - evolving and spanning a range from sub - luminous ( @xmath103 mag ) to luminous ( @xmath104 mag ) , accurately reflecting our expert knowledge applied in construction of the training set .
disagreement between object location and shaded labeling in figure [ fig : svm ] is partially due to variance in the observational data , which limits the efficiency of the classifier , and partially due to the coarse binning of the display slices .
we have configured the classifier with the penalty parameter @xmath105 and class weights calculated from their representation in the training set .
finally , we apply the classifier to the unclassified ( snii ? ) objects to infer their subclasses .
while `` sn iib '' and `` iin '' classifications can only be assigned by spectroscopy , by definition , here we are using photometric information to supplement available spectroscopy for the purpose of predicting the most likely spectroscopic sub - classification .
the classifier identifies 9 of these objects as sne iip , which we hereafter incorporate in our sn iip sample ; 6 as sne iib ; and 3 as sne iin .
when applied to the training set , the classifier achieves 83% accuracy across all sub - types , and 89% accuracy for sne iip .
it mistakenly classifies sne iib and iin as sne iip only 22% of the time ( the false identification rate ) , and mistakenly classifies sne iip as other classes only 11% of the time ( the false negative rate ) .
these statistics imply that , of our sample of 17 sne ii ?
, @xmath106 may be incorrectly classified as sne iip and @xmath106 legitimate sne iip may be inadvertently excluded . in combination with our sample of 67 securely - identified sne iip
, this level of false identification would imply a @xmath107 contamination rate within the sample . given the relatively low s / n of the sn ii ?
objects , the classifier may perform more poorly than on the training dataset ; but for the same reason , the influence of these objects on our light curve parameter inferences will be correspondingly suppressed .
in this section , we present the results of our analysis of the observational properties of the ps1 sn iip light curves .
the parameters calculated for individual objects are summarized in table [ tab : ppars ] .
we construct @xmath0 sn iip template light curves based on scaled stacks of our ps1 photometric sample .
figure [ fig : stacklc ] ( left ) shows these stacked sn light curves . in each band
, we only include objects with a @xmath108 uncertainty in the peak magnitude of @xmath109 mag and with an uncertainty in the plateau epoch of @xmath110 days , as estimated from the marginal posterior of the fitted light curve model . in total , we include @xmath111 $ ] objects in the @xmath112$]-band templates .
figure [ fig : stacklc ] ( right ) shows the template light curves constructed from these data .
the template magnitudes are calculated based on the range of observed photometry within a moving window of 6 rest frame days in width , between @xmath113@xmath114 days from peak magnitude .
template values are only reported if at least 4 photometric observations fall in the window . in total
, we include @xmath115 $ ] observations in constructing the @xmath112$]-band templates .
the templates are reproduced in table [ tab : templ ] .
past studies have varied widely in their interpretation of a putative , fast - declining sub - class of sne ii commonly labeled `` sne iil . ''
the sn iil classification was coined by @xcite , in recognition of a set of 6 type ii sne with an unusually fast decline rate of @xmath116 in the @xmath117-band , constituting @xmath118 of their sn ii sample . from a theoretical perspective ,
a fast declining sn ii suggests a less massive h envelope and consequently less sustained power contribution from h recombination .
likely explanations for diversity in h envelope mass among sn ii progenitors include variation in progenitor initial mass , with sne iil likely to arise from either progenitors with intrinsically less massive envelopes ( e.g. initial masses of @xmath119 ; @xcite ) or from more massive stars which have been stripped by strong radiation driven winds ( @xmath120 ; @xcite ) , and/or stars which have been partially stripped by interaction with binary companions @xcite .
observationally , no widely adopted criteria for sn iil classification have emerged .
@xcite divided their sn ii light curve set into `` linear '' and `` plateau '' classes based on the @xcite criterion , but they describe the decline rate distribution of these objects as continuous and find no statistical evidence for two modes in the distribution . in their population study of the loss sne , @xcite define sn iil as having sn iip - like spectroscopic features , but exhibiting an @xmath7-band decline of @xmath121 mag after explosion . by their selected criterion , they identify 7 unambiguous sne iil among 81 sne ii , and they determine sne iil to represent 10% of the volume - limited sn ii population .
they find that the sne iil are on average over - luminous compared to sne iip , and therefore would represent @xmath122 of a magnitude - limited sn ii survey sample . by inspection of a set of 22 sn ii light curves from the cccp and the literature ,
@xcite identify a set of 5 objects which they describe as sne iil .
their selection corresponds to a decline rate criteria of @xmath123 mag within 50 days of explosion .
they suggest that these objects form a class `` distinct '' from sne iip , and that the two classes together do not form a continuum of decline rates .
most recently , @xcite have characterized the decline rate distribution of hydrogen - rich sne as continuous .
we take an empirical approach to searching for evidence of a fast - declining mode in the sample of ps1 objects classified nominally as sne iip ( see section [ sec : mod : class ] ) .
figure [ fig : betakde ] shows the distribution of mean posterior values for the @xmath81 plateau decline rate parameter in our fitted light curve models , among objects with significant constraints on the plateau phase decay behavior .
we employ gaussian kernel density estimation ( kde ) to estimate the underlying form of the decline rate distribution from the observed data .
we select a kde bandwidth factor of 0.43 in @xmath61-band ( and similar in other bands ) based on the number of objects in the well constrained sample , following @xcite .
inspection of the kde models in figure [ fig : betakde ] does not indicate the presence of any fast - declining modes , e.g. within the shaded regions corresponding to the @xcite and @xcite criteria . given the limited sample size
, this test would not be sensitive to the existence of a small sn iil sub - population if they are subsumed within the fast - declining tail of a more numerous sn iip population .
however , a significant sub - population ( @xmath124 ) as proposed by previous authors should emerge if present .
we therefore conclude that empirical evidence for a distinct sn iil sub - population is not present within the ps1 sn ii dataset , and hereafter we do not distinguish between sne iil and iip . discounting the existence of an sn iil subclass differs from the interpretation applied in many past studies , but does not disagree quantitatively with past results .
@xcite adopted an sn iil classification scheme defined arbitrarily by previous authors @xcite for the purpose of estimating volumetric rates .
@xcite applied a similar criterion , then tested for a discontinuity between the properties of this sub - population and the `` normal '' sn iip population .
they reported a negative result .
@xcite identified a slow declining subset of objects in their sample , but did not quantitatively test for a discontinuity .
moreover , we note that our sample , consisting of 112 sne ii in total , is @xmath125 larger than the @xcite sn ii sample and @xmath126 times larger than the @xcite sample . using a similarly sized sample ,
@xcite have reported independent results in agreement with the continuum of decline rates we report here .
figure [ fig : stacklc ] further illustrates this result , showing our sn iip @xmath61-band photometry in comparison to the averaged unfiltered sn iip and sn iil light curves of @xcite .
both their templates fall within the photometric range of our observed sample , with their sn iip template falling near the @xmath127 range of our light curves and their sn iil template falling near the median of our light curves .
note that their photometry was collected with an unfiltered photometric system , which @xcite compare to @xmath7 band .
this would suggest that their response function includes redder wavelengths than our @xmath61-band photometry .
this may partially explain the slower decline rates in their sample , as redder filters ( e.g. @xmath128-bands ) exhibit more gradual sn iip decline rates in the plateau phase .
regardless , this comparison illustrates that the long - recognized range of decline rates among sne with broad h@xmath26 features falls within the observed , continuous range of decline behavior exhibited by objects classified as sne iip in our sample , with no evidence for a second , fast - declining mode in the population .
recent observational studies have suggested that the plateau durations of sne iip are tightly distributed around @xmath6 days @xcite , in stark contrast to theoretical predictions that they should vary from @xmath129 days given the expected range of progenitor properties including mass , energy , and radius ( see e.g. @xcite ) .
@xcite has interpreted this as evidence that the joint parameter space of progenitor properties must be tightly constrained , with the explosion energy scaling precisely with the cube of the mass to produce a roughly constant plateau duration .
we quantitatively assess the plateau duration distribution of our sne iip sample . in agreement with @xcite , @xcite , and @xcite , the @xmath61-band plateau duration distribution of our sample ( figure [ fig : tpfdist ] )
is peaked around @xmath130 days , with median value of 92 days and @xmath108 variation of 14 days .
however , the full observed range of plateau durations is significantly larger , @xmath131 days . note that we do not include in this distribution poorly - constrained light curves , where the marginalized posterior uncertainty is @xmath132 of the prior variance ( @xmath133 days at @xmath108 ) . for those poorly constrained objects , as figure [ fig : tpfdist ] illustrates ,
the posterior is typically peaked around the mean of the prior distribution .
our ps1 sample therefore suggests that the observed variation in sn iip plateau durations is not as highly constrained as suggested by @xcite and @xcite .
however , we do not detect objects with plateaus as long lived as the least energetic objects produced in the model grid of @xcite ( e.g. @xmath134 days for kinetic explosion energy of @xmath135 at infinity ) .
however , all wide field surveys will be insensitive to identifying the longest duration objects for two primary reasons .
first , such low energy explosions will have correspondingly low luminosities ( @xmath136 , similar to the least luminous objects in our sample ; see table [ tab : ppars ] ) and therefore be subject to malmquist bias .
second , because the length of the observing seasons of each ps1 md field ( typically @xmath137 days ) are similar to the sn iip plateau duration , there will be an observational bias towards measuring well - constrained plateau durations only for short - duration objects .
we therefore interpret the results presented here as a lower limit on the true variation in the sn iip plateau duration distribution . in a companion paper , @xcite
, we use a hierarchical bayesian light curve fitting methodology to assess the influence of these biases on our inferred plateau duration distribution , and to estimate the underlying plateau duration distribution .
we correct for total extinction ( galactic and host environment ) using the @xmath39 color excess method of @xcite as revised by @xcite see figure [ fig : viext ] .
we calculate @xmath138 by comparing to the theoretical baseline @xmath39 color photometry to the landolt and vega systems using a color - based @xmath139 correction obtained by integrating over the temporally nearest available uvoir spectrum of sn 1999em , the @xmath140 day spectrum from @xcite .
we find that @xmath141-corrections based on the spectrophotometric templates applied in section [ ssec : kcor ] would introduce a systematic discrepancy in the @xmath142 color relative to sn iip corrections in past works ( e.g. @xcite ) of @xmath143 mag , due to differences in the equivalent width of the nir ca ii emission feature .
we estimate a @xmath64 mag systematic uncertainty may remain in our @xmath144-band magnitude estimates due to variation in this feature .
we apply the ab offset of @xcite . ] for the given @xmath145-band luminosity ( equation 15 of @xcite ) , and convert to extinction in each @xmath13 band using the extinction curve of @xcite and assuming @xmath146 .
we apply a physically - motivated uniform prior enforcing @xmath147 mag .
the median uncertainty for the extinction estimate of an individual sn is 0.4 mag .
we find extinctions varying from @xmath148 mag , with median value of @xmath149 mag . in this distribution
we have excluded objects with extinction uncertainties @xmath150 mag , affecting 8 objects .
this median is somewhat higher than the mean extinction reported by @xcite for their sample of sne iip with progenitor imaging , @xmath151 mag .
the discrepancy may result from a combination of differences in methodology ( the estimates from @xcite are derived from a heterogeneous set of sources and techniques ) and sample selection ( given that our ps1 transient search operates in red bands , particularly @xmath52 and @xmath41 , it may be relatively insensitive to selection against high extinction objects ) .
both observations and theory show that there is significant scatter around the fiducial @xmath39 color , indicating intrinsic color variation not accounted for in our color excess - based extinction estimate .
moreover , the color excess method may be particularly vulnerable to under estimates for low - metallicity progenitors where decreased iron line blanketing would reduce suppression of blue flux .
we note that @xcite have recently shown that both photometric and spectroscopic sn iip extinction correction prescriptions have limited effectiveness ( as measured by their ability to reduce the scatter in observed colors ) , and exhibit systematic offsets relative to each other .
we show the peak absolute magnitude distribution for our ps1 sn iip sample in each photometric filter in figure [ fig : mpeakdist ] .
we correct for extinction using the @xmath39 color excess method , as described in section [ sec : prog : ext ] .
statistics of these distributions are given in table [ tab : absmagdist ] . in this table and hereafter , we include only objects with well constrained posterior distributions : @xmath152 mag at @xmath108 without extinction correction , and @xmath153 mag with correction .
the median uncertainty for the absolute magnitude measurement ( @xmath108 posterior width ) is @xmath154 of @xmath155 mag in @xmath61 band without extinction correction , and @xmath156 mag after correction .
the observed peak magnitude distribution agrees well with the @xmath7 band luminosity function of @xcite , reflected in the prior distribution shown in figure [ fig : mpeakdist ] ( see also section [ sec : mod : fit ] and figure [ fig : mpprior ] ) .
the comparison is somewhat limited , as @xcite examined a different filter ( @xmath7 ) and did not correct for extinction ( although the objects in their sample should have low extinction ) .
note that this similarity suggests agreement of the observed distribution of peak magnitudes between our surveys .
if the prior were simply dominating our inferred peak magnitudes , the posterior medians would all be centered at the prior mean ( @xmath157 ) .
this effect can be seen in the @xmath15-band peak magnitude distribution shown in figure [ fig : mpeakdist ] , where prior information does dominate for a substantial fraction of objects .
extinction correction shifts the median of the distribution by @xmath158 mag in @xmath47-band , and only @xmath64 mag in @xmath41-band .
we regard the @xmath15-band distribution as least reliable due to the small number of objects ( @xmath159 ) that pass our posterior width cut , and therefore do not include them in table [ tab : absmagdist ] .
the extinction corrected peak magnitude distribution spans a @xmath108 range of @xmath160 to @xmath161 mag in @xmath61-band , indicating a factor of @xmath106 spread in the luminosity distribution of these sne .
this wide range in sn properties has important implications for the mass distribution of their progenitor stars , which we will model and discuss in section [ sec : prog : mass ] .
@xcite found a @xmath117-band volume limited peak magnitude distribution with mean and standard deviation of @xmath162 among 74 sne iip from the literature , @xmath163 mag dimmer than the @xmath47-band mean peak magnitude we report .
the discrepancy in the mean of the absolute magnitude distributions between our study and theirs is likely due to a combination of sample selection effects and extinction correction methodology . @xcite
have considered sne iil separately from sne iip , which systematically lowers their sn iip peak magnitude distribution relative to ours ( see section [ sec : res : iil ] ) .
morever , @xcite have applied a statistical correction for extinction , assuming a mean value of @xmath164 mag for each core - collapse sn , which is typically @xmath165 mag lower than the extinction correction we measure among our sne iip with the @xmath39 color excess method ( section [ sec : prog : ext ] ) .
llrrrlrrr @xmath47 & 36 & -16.86 & -17.86 & -18.43 & 13 & -17.84 & -19.14 & -19.70 + @xmath61 & 50 & -16.92 & -17.80 & -18.45 & 23 & -17.47 & -18.10 & -19.28 + @xmath52 & 49 & -17.08 & -17.68 & -18.33 & 26 & -17.31 & -18.10 & -18.94 + @xmath41 & 40 & -17.11 & -17.89 & -18.42 & 33 & -17.63 & -18.06 & -19.25 upon investigation of the joint posterior parameter distributions among our fitted sn iip light curves , we identified a highly significant relation between the plateau phase decline rate ( @xmath81 ) and the extinction corrected peak magnitude ( @xmath166 ) of the sne .
figure [ fig : phasemprbeta2 ] illustrates that these parameters are highly correlated in @xmath61-band , with pearson correlation coefficient of @xmath167 ( @xmath168-value of @xmath169 ) .
we have evaluated the strength of the correlation between these parameters in all combinations of photometric filters available from the ps1 dataset ( figure [ fig : phasemprbeta2bands ] ) and identified this relation to be significant in every band ( except @xmath15 , where the ps1 photometry is limited ) , though it is strongest in the @xmath61 and @xmath52-bands . in the @xmath61-band ,
the relation between these parameters is best fit with the relation : @xmath170 & = ( { -13.1}\pm { 1.2 } ) + m_{peak}[r ] ~ ( { -0.47}\pm { 0.07})\end{aligned}\ ] ] this linear fit was obtained using a bayesian methodology accounting for the two dimensional covariance of the @xmath166 and @xmath81 posterior distributions for each sn , and modeling the relationship between these values with an intrinsic scatter @xmath145 orthogonal to the axis of the fit line @xcite .
the @xmath171$]th percentile values of the intrinsic scatter parameter , projected to the peak magnitude axis , are @xmath172 $ ] mag .
no samples among the @xmath173 returned from our mcmc fit of this model have positive slopes for the linear fit , supporting the low @xmath168-value reported from the pearson correlation coefficient test above .
when we perform the same analysis on the observed absolute peak magnitudes , not corrected for extinction , we find a similar trend , but with somewhat decreased significance ( @xmath168-value of @xmath174 and projected intrinsic scatter median @xmath175 mag ) .
this result suggests that application of the color - excess based extinction estimates provides measurable success in recovering the intrinsic distribution of light curve properties , despite the shortcomings described in section [ sec : prog : ext ] .
considering the @xmath145-band data of @xcite parameters to our @xmath81 given that @xmath176 . ] and applying the same bayesian linear fitting methodology yields a median intrinsic scatter of @xmath177 mag . the larger intrinsic scatter indicated by the @xcite data results in part from the much smaller uncertainties in the light curve parameters of the nearby sne they study .
this discrepancy may imply that the uncertainties in our light curve parameters are somewhat overestimated , or that theirs are somewhat underestimated . in past studies , divisions of sn iip samples into `` normal '' sn iip and sn iil ( see section [ sec : res : iil ] )
sub - classes may have obscured this relation by reducing the dynamic range available for correlations against the decline rate .
@xcite have recently reported an independent discovery of this relation using a similarly - sized sample of @xmath145-band light curves from a variety of past sn searches .
this independent discovery , and the multi - band manifestation of the relation that we explore here , underscores its physical significance .
this result echoes past conclusions that `` sne iil '' are on average more luminous than more slowly declining sne iip ( e.g. @xcite ) . the existence of a decline rate
peak magnitude relation for sne iip is reminiscent of the well known width
luminosity relation for sne ia ( e.g. * ? ? ?
however , while brighter sne ia have more slowly declining light curves , brighter sne iip instead have faster declining light curves ( figure [ fig : phasemprbeta2 ] ) .
moreover , the sn iip relation has significant intrinsic scatter : @xmath178 mag . that the sn iip decline rate
peak magnitude relation is recoverable at all is testament to the wide range of variation in their plateau phase decline rates , which ranges a factor of @xmath179 . the sn ia decline rate
peak magnitude relation is interpreted as evidence for a fundamental plane or single - parameter family among sne ia light curves , and a similarly sparse dimensionality may apply to sne iip .
several authors have argued for a single parameter source of variation among sne iip light curves @xcite , particularly with respect to the luminosity velocity relation applied in the standardizable candle method @xcite . interpreted together with the results of section [ sec : res : iil ] , this would seem to point to the explosion energy , likely determined by the progenitor initial mass through its influence on the mass of the hydrogen envelope at the time of explosion , as the predominant single parameter continuously determining the fundamental observational properties of the explosions of hydrogen - rich red supergiants , as @xcite also concluded .
however , the presence of significant scatter around the sn iip decline rate
peak magnitude relation presented here suggests additional correlations with secondary factors , as have emerged among sne ia @xcite .
relevant secondary characteristics of the progenitor stars are likely to include metallicity and rotation velocity . in general , more massive h - rich envelopes should have longer radiative diffusion times ( @xmath180 ) , suggesting slower light curve evolution and lower peak magnitudes .
however , @xcite have found an empirical linear relation between the progenitor initial mass of sne iip and their expansion velocities in the plateau phase . with these two factors in opposition , this would suggest that the diffusion time , and perhaps the @xmath81 parameter , should be nearly constant among sne iip
. this discrepancy could result from nonlinearity in the relationship between progenitor initial mass and the mass of the hydrogen envelope at the time of explosion or variation in the velocity evolution between sne iip .
alternatively , it may challenge the robustness of the linear mass velocity relation constructed by @xcite .
we compare the light curve models for each sn iip in our sample to relevant theoretical models to estimate the physical properties of their progenitor stars .
[ sec : res : boll ] for the purpose of comparing to the full - wavelength spectrophotometric models of @xcite , we estimate the pseudo - bolometric ( oir ) luminosity of each sn by integrating over the multi - band ( @xmath0 ) light curve model described in section [ sec : mod : fit ] .
we apply a correction to the pseudo - bolometric light curves to , primarily , account for unobserved ir flux and some uv emission .
we calculate this correction ( @xmath181 ) by integrating a sn iip spectral template from a similar epoch ( @xmath182 days ; @xcite ) over the @xmath183 bands , obtaining @xmath184 mag . by using a fixed spectral template
, we neglect any intrinsic variation in the sn iip spectral properties .
the modeling in @xcite suggests that any intrinsic color variation among sne iip is small , with the bolometric correction estimated to vary by @xmath64 mag based on differences in progenitor metallicity and initial mass ( their figure 14 ) .
we estimate the ejected mass of radioactive @xmath84ni ( @xmath185 ) by comparison to the late - time light curve of sn 1987a ( see figure [ fig : qbolo ] ) , following the method of @xcite .
we compare our pseudo - bolometric light curves ( section [ sec : res : boll ] ) to the sn 1987a light curve using the late - time ( @xmath186 days ) rest frame epoch where the model is most tightly constrained and assuming complete gamma - ray trapping , as in @xcite ; but see also the discussion of gamma - ray trapping in @xcite .
the distribution of @xmath185 among our ps1 sne iip has @xmath171$]th percentile values of @xmath187~\rm{m}_\odot$ ] ( see figure [ fig : mnidist ] ) .
we include in this distribution only objects with photometry to directly constrain their light curve evolution in the radioactive decay dominated phase .
we identify such objects by looking at the posterior variance in the @xmath82 parameter , which is @xmath188 dex at @xmath108 in cases where the prior alone sets the parameter value .
we therefore adopt a cut at @xmath189 dex to exclude poorly constrained ni masses .
we note that , among the entire sample , the typical uncertainty in @xmath185 is fairly large ( median of @xmath190 dex ) due to the relatively poor photometric coverage of the late phases of sne in the sample . among the excluded objects , some have unphysically large median estimates for the ejected @xmath84ni mass ( @xmath191 ) ;
the high uncertainty in these estimates is appropriately reflected in the confidence intervals quoted in table [ tab : ppars ] .
our results suggest a factor of @xmath192 smaller @xmath185 for most sne iip in comparison to typical sne ib and ic , which have @xmath193 @xcite .
this is consistent with the expectation that sne ib and ic are produced by more massive stellar progenitors than sne iip .
our measured @xmath185 distribution is similar to that reported by @xcite , who derive @xmath185 ranging from @xmath194 for a set of 11 sne iip from the literature , with a mean value of @xmath195 .
we derive physical parameters ( including progenitor initial mass , radius , and explosion energy ) for the progenitor stars of our sn iip sample by comparison to the modeling of @xcite . in general ,
more massive progenitor stars produce plateaus that are brighter and of longer duration .
there is a secondary contribution from @xmath185 which has the effect of increasing the plateau duration . @xcite
provide scaling relations for the observable parameters @xmath196 ( bolometric luminosity at @xmath197 days after explosion ) and @xmath198 ( plateau phase duration , corrected for extension by ejected @xmath84ni ) as a function of @xmath199 ( total explosion energy in units of @xmath200 ergs ) and @xmath2 ( initial mass of progenitor star ) .
we note that there is a well known discrepancy between masses derived from hydrodynamic light curve modeling and direct progenitor searches , with progenitor searches yielding lower mass estimates @xcite .
we derive @xmath198 using their equation 13 , which requires an estimate for @xmath185 ( see section [ sec : res : decay ] ) , @xmath201 ( the progenitor radius at the time of explosion ) and @xmath202 ( the total ejected mass ) . following the recommendations of @xcite
, we calculate values for @xmath201 and @xmath202 strictly as a function of @xmath2 by interpolating over the progenitor star model grid used in their study . solving this system of equations for every sample in the posterior distribution of the light curve model , we obtain estimates for @xmath2 , @xmath202 , and @xmath199 for each sn iip in our sample . in figure
[ fig : ppar ] , we compare the observed plateau duration and @xmath196 to model curves from @xcite .
this comparison illustrates that the observed variation in sn iip light curve phenomenology is significantly larger than the range explored by the @xcite model grid , preventing us from performing progenitor parameter estimation for a substantial number of the objects in our sample .
given the observed @xmath185 values measured for each individual light curve , 67 of our sne iip have light curve properties outside the region covered by the @xcite models , and we can produce mathematically valid comparisons to model light curves from the @xcite grid for only 8 sne iip ( see table [ tab : ppars ] ) .
even these comparisons typically yield explosion energies at or above the most energetic models in their grid ( @xmath203 ) .
a large number of objects in our sample have luminosities lower than the range of theoretical values for a given plateau duration ( falling in the blue shaded region of figure [ fig : ppar ] ) ; accounting for 61 of the 67 objects for which we do not report progenitor inferences .
we therefore advocate for further theoretical exploration of the sn iip light curve parameter phase space through additional hydrodynamic modeling , in order to provide more robust model grids and scaling relations for comparison to observational surveys .
in particular , certain modeling systematics should be addressed to enable inference on objects occupying the low - luminosity and short plateau duration parameter space .
@xcite have previously recognized a systematic discrepancy between sn iip model and observed light curves , and identified over - sized radii in model progenitor stars as the likely cause .
hydrodynamic explosion simulations based on more advanced progenitor star models , taking into account important effects such as rotation on the stellar radius ( e.g. @xcite ) , are needed to overcome this bias . moreover , synthetic light curves for sne with progenitor initial masses @xmath204 , down to the limit for core - collapse ( e.g. @xmath205 , see @xcite and references therein ) , and higher explosion energies should be produced .
such low mass progenitors have already been explored in the context of extremely low luminosity sn iip like sn 1997d @xcite , and model light curves of this kind calculated with full self - consistency across the mass and energy spectrum are needed .
additionally , other progenitor physical parameters will play important secondary roles in determining sn light curve properties and should be included in future simulations . in particular , lower metallicity can serve to lower the plateau luminosity by decreasing the opacity of the h envelope , therefore lowering plateau luminosities , and by decreasing mass loss , causing longer lived h recombination and plateau durations @xcite .
@xcite have recently performed a theoretical survey of sn iip radiative properties using a grid of synthetic progenitor stars expanded in parameter scope with respect to the grid of @xcite , including core overshoot , rotation , mixing , and metallicity .
however , the @xcite grid was limited to a single initial progenitor mass of @xmath206 .
an expanded survey of the kind performed by @xcite , to provide updated observationally - applicable scaling relations of the kind provided by @xcite , would be an invaluable tool for inference on the progenitor population of massive star explosions in the nascent untargeted survey era .
photometric galaxy metallicity diagnostics like that presented in @xcite can be used to observationally estimate metallicities for the progenitor stars of observed sne to facilitate model comparison .
while theoretical stellar evolution modeling has historically predicted that stars undergoing core - collapse up to @xmath207 should produce type iip sne ( see e.g. @xcite and references therein ) , a variety of observational techniques have failed to identify sn iip progenitors with masses greater than @xmath208 .
the shape of the stellar initial mass function can at least help to explain this absence , with @xmath209 progenitors being a factor of @xmath210 times more common than @xmath207 progenitors .
however , even accounting for imf prescriptions of varying steepness , @xcite reported an upper limit for the sn iip progenitor mass of @xmath211 based on their direct progenitor detection search .
similarly , in an x - ray analysis of sn iip progenitor mass loss rates , @xcite estimated a maximum mass loss rate of @xmath212 .
black hole fallback formation @xcite and / or progenitor rotation @xcite may provide a theoretical solution to this `` red supergiant problem . ''
our results have only indirect implications for the red supergiant problem .
while our progenitor mass estimates ( e.g. figure [ fig : ppar ] ) show no evidence for an absence of sn iip progenitor stars at the high mass end of the @xcite progenitor grid , our mass estimates are only as reliable as the calibration of the hydrodynamic modeling underpinning them .
as we have discussed , there are known systematic offsets between the masses inferred from hydrodynamic modeling and progenitor star photometry , and additional theoretical investigation of this discrepancy is needed to conclusively resolve the red supergiant problem .
however , our observations and modeling suggest that there is no discontinuous distinction between type iip and iil sne ( section [ sec : res : iil ] ) . given the correlation between plateau phase decline rate and luminosity ( section [ sec : res : pcor ] ) , and the theoretical connection between luminosity and progenitor mass , the exclusion of luminous `` sne iil '' could obscure the high - mass end of the sne iip progenitor population . emphasizing the role of small number statistics in this regime , we note that @xcite report a progenitor mass upper limit that is already consistent with @xmath213 at @xmath214 , and consistent with @xmath215 at @xmath214 if only events with robust progenitor detections are included in the analysis .
@xcite have recently considered the extinction due to the dust produced by red supergiant winds and revised the sn iip upper mass limit to @xmath216 ( with @xmath214 uncertainty ) .
as the direct progenitor search technique has so far only been applied to a handful of objects , we look to future detections of the progenitors of luminous sne iip to address the red supergiant problem ( see e.g. @xcite ) .
we have assembled and studied the full sample of type iip sn light curves from the medium deep survey , totaling 18,953 photometric data points ( 5,096 robust transient detections ) in @xmath0 filters for 112 individual sne ( 76 sne iip ) .
we have developed a bayesian light curve fitting methodology for sne ii based on a physically - motivated 5-component segmentation of the sn iip light curve ( section [ ssec : mod ] ) .
we present an implementation of our sn iip model for use with the hamiltonian monte carlo library _ stan _ in appendix [ ap : stan ] .
we have interpreted our light curve modeling in terms of the hydrodynamic type iip sn progenitor model grid of @xcite .
the primary conclusions of this work are as follows : * we present photometric @xmath27-corrections and sn iip light curve templates in the @xmath0 bands ( sections [ ssec : kcor ] and [ sec : res : stack ] ) .
our templates are based on 2,804 individual photometric detections for 66 individual sne iip from @xmath217 to @xmath218 days from peak magnitude . *
consistent with theoretical expectations , our sn iip sample spans a diverse range of light curve parameters : a factor of @xmath219 in plateau phase decay rate ( figure [ fig : betakde ] ) , @xmath192 orders of magnitude in ejected @xmath84ni ( figure [ fig : mnidist ] ) , a factor of @xmath192 in plateau duration ( figure [ fig : tpfdist ] ) , and @xmath220 mag in absolute magnitude ( figure [ fig : mpeakdist ] ) .
this evidence stands in contrast to recent suggestions by @xcite and @xcite that the sn iip plateau duration distribution , a critical observational parameter tied to progenitor initial mass , is tightly distributed ( section [ sec : res : tp ] ) . *
addressing a longstanding debate in the literature , we have searched for the existence of a fast declining `` sn iil '' sub - population in the decline rate distribution of our sn iip sample ( section [ sec : res : iil ] ) .
we find no evidence for a discontinuity in this distribution for any photometric band , questioning the existence of a discrete sn iil sub - population .
* we identify a highly significant statistical correlation between the peak magnitude and plateau phase decline rate of sne iip ( section [ sec : res : pcor ] ) . together with the previous results ,
this supports the interpretation of core - collapse among hydrogen - rich red supergiants as a predominantly single parameter family of explosions , whose observational behavior is determined primarily by the explosion energy and likely set by the initial mass of the progenitor star .
this represents an independent discovery and confirmation of results recently reported in @xcite .
* through the largest systematic comparison to date of sn iip lightcurves to hydrodynamic progenitor models , we have derived mass , radius , and explosion energy estimates for the objects in our sample ( section [ sec : prog : mass ] ) . however , we find that the available theoretical model grids are insufficient to cover the full range of observed variation in sn iip light curve properties .
we point to the need for additional hydrodynamic modeling to produce updated and expanded self - consistent model grids , particularly in the low luminosity regime . *
though our progenitor inferences are based on hydrodynamic light curve models , which are known to produce systematically higher masses than direct progenitor detection searches , we do not find evidence for an absence of high - mass sn iip progenitors ( section [ sec : prog : rsgprob ] ) .
we point to future direct progenitor detections of luminous sne iip as having the potential to ease the discrepancy between the maximum sn iip progenitor mass identified by various theoretical and observational methods , known as the red supergiant problem . in a companion paper , @xcite
, we discuss and demonstrate a hierarchical expansion of the model presented here to provide a general framework for the analysis of supernova light curve populations in the coming era of next generation wide field transient searches .
we advocate continued investment in statistical and computational tools in the future as a means to compensate for the relative decline anticipated in the availability of detailed spectroscopic and other follow - up information on individual transients .
[ sec : ackn ] we thank k. mandel for sage guidance and many helpful conversations ; d. kasen for his thoughtful comments on this work ; an anonymous referee for their very helpful review ; and the _ stan _ team for their excellent modeling language and hmc sampler .
the pan - starrs1 surveys ( ps1 ) have been made possible through contributions of the institute for astronomy , the university of hawaii , the pan - starrs project office , the max - planck society and its participating institutes , the max planck institute for astronomy , heidelberg and the max planck institute for extraterrestrial physics , garching , the johns hopkins university , durham university , the university of edinburgh , queen s university belfast , the harvard - smithsonian center for astrophysics , the las cumbres observatory global telescope network incorporated , the national central university of taiwan , the space telescope science institute , the national aeronautics and space administration under grant no .
nnx08ar22 g issued through the planetary science division of the nasa science mission directorate , the national science foundation under grant no .
ast-1238877 , the university of maryland , and eotvos lorand university ( elte ) .
observations reported here were obtained at the mmt observatory , a joint facility of the smithsonian institution and the university of arizona .
this paper includes data gathered with the 6.5 meter magellan telescopes located at las campanas observatory , chile .
based on observations obtained at the gemini observatory , which is operated by the association of universities for research in astronomy , inc .
, under a cooperative agreement with the nsf on behalf of the gemini partnership : the national science foundation ( united states ) , the national research council ( canada ) , conicyt ( chile ) , the australian research council ( australia ) , ministrio da cincia , tecnologia e inovao ( brazil ) and ministerio de ciencia , tecnologa e innovacin productiva ( argentina ) .
the data presented here were obtained in part with alfosc , which is provided by the instituto de astrofisica de andalucia ( iaa ) under a joint agreement with the university of copenhagen and notsa .
support for this work was provided by the david and lucile packard foundation fellowship for science and engineering awarded to a.m.s .
is supported under epsrc grant ep / j016934/1 _ facilities : _ 100 natexlab#1#1 , h. , et al .
2011 , , 193 , 29 , j. p. , et al .
2014 , arxiv e - prints , 1403.7091 , i. , et al .
2012 , , 756 , l30 , w. d. 1980 , , 237 , 541 , r. , ciatti , f. , & rosino , l. 1979 , , 72 , 287 , e. , nugent , p. e. , branch , d. , & hauschildt , p. h. 2004 , , 616 , l91 , m. c. , & hamuy , m. 2009 , , 701 , 200 , m. r. , & roweis , s. 2007 , , 133 , 734 , j. a. , clayton , g. c. , & mathis , j. s. 1989 , , 345 , 245 chambers , k. c , _
et al._. in preparation , a. , domnguez , i. , hflich , p. , limongi , m. , & straniero , o. 2003 , , 345 , 111 , n. n. , & utrobin , v. p. 2000 , , 354 , 557 , d. h. , mccrea , w. h. , & stephenson , f. r. 1977 , , 265 , 318 , c. b. , et al .
2010 , , 708 , 661 , l. , & hillier , d. j. 2011 , , 410 , 1739 , l. , hillier , d. j. , waldman , r. , & livne , e. 2013 , , 433 , 1745 , a. , hare , t. , bigelow , b. c. , & osip , d. j. 2006 , in society of photo - optical instrumentation engineers ( spie ) conference series , vol .
6269 , society of photo - optical instrumentation engineers ( spie ) conference series , m. r. , et al .
2011 , , 741 , 97 , v. v. 2014 , arxiv e - prints , 1402.5150 , s. , et al .
2012 , , 537 , a146 , j. j. 2008 , royal society of london philosophical transactions series a , 366 , 4441 , n. , et al .
2011 , , 742 , 6 , d. , et al .
2005 , , 117 , 1411 , t. , et al .
2014 , arxiv e - prints , 1404.0378 , j. , mahabal , a. , sun , j. , wang , x. , yi , wang , & zhang , l. 2014 , arxiv e - prints , 1401.3211 , a. v. 1997 , , 35 , 309 , r. j. , sanders , n. e. , & kirshner , r. p. 2011
, , 742 , 89 , m. , et al .
2010 , , 714 , l280 , c. l. 1999 , , 522 , 413 gelman , a. , carlin , j. , stern , h. , dunson , d. , vehtari , a. , & rubin , d. 2013 , bayesian data analysis , third edition , chapman & hall / crc texts in statistical science ( taylor & francis ) gelman , a. , jakulin , a. , pittau , m. g. , & su , y .- s .
2008 , the annals of applied statistics , 2 , 1360 , c. , et al .
2013 , , 558 , a103 , s. , et al .
2010 , , 720 , l77 , r. l. , nugent , p. e. , & phillips , m. m. 1999 , , 521 , 30 , j. h. , meynet , g. , & ekstrm , s. 2013 , , 550 , l7 , j. h. , meynet , g. , georgy , c. , & ekstrm , s. 2013 , , 558 , a131 , m. 2003 , , 582 , 905 , m. , & pinto , p. a. 2002 , , 566 , l63 , a. , fryer , c. l. , woosley , s. e. , langer , n. , & hartmann , d. h. 2003 , , 591 , 288 , k. w. , siegmund , w. a. , kaiser , n. , chambers , k. c. , laux , u. , morgan , j. , & mannery , e. 2004 , in society of photo - optical instrumentation engineers ( spie ) conference series , vol . 5489 ,
society of photo - optical instrumentation engineers ( spie ) conference series , ed .
j. m. oschmann jr . , 667678 hoffman , m. d. , & gelman , a. 2014 , journal of machine learning research , 15 , 1593 , d. w. , baldry , i. k. , blanton , m. r. , & eisenstein , d. j. 2002 , arxiv astrophysics e - prints , d. w. , bovy , j. , & lang , d. 2010 , arxiv e - prints , 1008.4686 , i. m. , jrgensen , i. , allington - smith , j. r. , davies , r. l. , metcalfe , n. , murowinski , r. g. , & crampton , d. 2004 , , 116 , 425 , a. , smartt , s. j. , fraser , m. , fransson , c. , sollerman , j. , taddia , f. , & kotak , r. 2013 , arxiv e - prints , 1311.2031 , n. , et al .
2002 , in society of photo - optical instrumentation engineers ( spie ) conference series , vol . 4836 , society of photo - optical instrumentation engineers ( spie ) conference series , ed . j. a. tyson & s. wolff , 154164 , d. , & woosley , s. e. 2009 , , 703 , 2205 , e. , et al .
2011 , , 192 , 18 , c. , robert , c. , & drissen , l. 1992 , , 401 , 596 , d. c. , et al .
2002 , , 114 , 35 , w. , et al .
2011 , , 412 , 1441 , w. , wang , x. , van dyk , s. d. , cuillandre , j .- c . ,
foley , r. j. , & filippenko , a. v. 2007 , , 661 , 1013 , c. h. , fenner , y. , & gibson , b. k. 2004 , science , 303 , 59 , i. y. , & nadezhin , d. k. 1985 , soviet astronomy letters , 11 , 145 , e. 2006 , in the advanced maui optical and space surveillance technologies conference , e. a. , liu , m. , monet , d. g. , & chambers , k. c. 2008 , in iau symposium , vol .
248 , iau symposium , ed .
w. j. jin , i. platais , & m. a. c. perryman , 553559 , k. , et al .
2010 , , 404 , 981 . 2012 , , 420 , 3451 , g. , eggenberger , p. , ekstrom , s. , georgy , c. , groh , j. , maeder , a. , saio , h. , & moriya , t. 2013 , arxiv e - prints , 1308.5797 , g. , et al .
2007 , , 666 , 674 , d. , et al .
2013 , , 767 , 71 , d. k. 2003 , , 346 , 97 , e. , & sari , r. 2010 , , 725 , 904 , k. , kobayashi , c. , & tominaga , n. 2013 , , 51 , 457 , k. , tominaga , n. , umeda , h. , kobayashi , c. , & maeda , k. 2006 , nuclear physics a , 777 , 424 , k. i. , iwamoto , k. , & suzuki , t. 1995 , , 256 , 173 , p. , et al .
2006 , , 645 , 841 , f. 2008 , arxiv e - prints , 0810.5518 , a. , et al .
2004 , , 347 , 74 , f. , barbon , r. , cappellaro , e. , & turatto , m. 1993 , , 98 , 443 . 1994 , , 282 , 731 pedregosa , f. , et al .
2011 , journal of machine learning research , 12 , 2825 , m. m. , lira , p. , suntzeff , n. b. , schommer , r. a. , hamuy , m. , & maza , j. 1999 , , 118 , 1766 , d. 2013 , , d. , et al .
2009 , , 694 , 1067 , i. , & waxman , e. 2011 , , 728 , 63 , a. , et al .
2013 , arxiv e - prints , 1310.3828 . 2005 , , 634 , 1103 , d. , jenkins , iii , r. l. , wright , j. , & maddox , l. 2014 , arxiv e - prints , 1403.5755 , n. , betancourt , m. , & soderberg , a. 2014 , arxiv e - prints , 1404.3619 , n. e. , levesque , e. m. , & soderberg , a. m. 2013 , , 775 , 125 , g. d. , weymann , r. j. , & foltz , c. b. 1989 , , 101 , 713 scott , d. 1992 , multivariate density estimation : theory , practice , and visualization , a wiley - interscience publication ( wiley ) , s. j. 2009 , , 47 , 63 , s. j. , eldridge , j. j. , crockett , r. m. , & maund , j. r. 2009 , , 395 , 1409 , n. , li , w. , filippenko , a. v. , & chornock , r. 2011 , , 412 , 1522 . 2014 ,
stan : a c++ library for probability and sampling , version 2.2 . 2014 , stan modeling language users guide and reference manual , version 2.2 , a. m. , dalcanton , j. j. , warren , s. r. , weisz , d. r. , skillman , e. , ott , j. , williams , b. f. , & dolphin , a. e. 2013 , , 772 , 124 , c. w. , doherty , p. , cramer , c. , narayan , g. , brown , y. j. , lykke , k. r. , woodward , j. t. , & tonry , j. l. 2010 , , 191 , 376 , d. a. , wheeler , j. c. , & harkness , r. p. 1991
, , 374 , 266 , k. , et al .
2013 , arxiv e - prints , 1311.2525 , f. x. , woosley , s. e. , & weaver , t. a. 1995 , , 98 , 617 , j. , & onaka , p. 2009 , in advanced maui optical and space surveillance technologies conference , , j. l. , et al .
2012 , , 750 , 99 , v. p. , & chugai , n. n. 2008 , , 491 , 507 , j. j. , & eldridge , j. j. 2012 , , 419 , 2054 , x. , et al .
2009 , , 699 , l139 , d. g. , et al .
2000 , , 120 , 1579 , t. r. 2004 , , 617 , 1233
below we reproduce the full bayesian model for our 5-component segmented sn ii light curve model described in section [ sec : mod : fit ] , in the _ stan _ modeling language .
the stan model specification format is documented in the stan modeling language users guide and reference manual @xcite .
the model takes the following data as input : ` n_obs ` , the total number of photometric data points ; ` n_filt ` , the number of photometric filters ; ` t ` , a vector of mjd dates of the photometric observations ; ` fl ` , a vector of luminosities corresponding to the photometric observations ( with units as described in section [ ssec : mod ] ) ; ` dfl ` a corresponding vector of luminosity uncertainties ; ` z ` the redshift ; ` t0_mean ` an initial estimate of the explosion date ( for initialization and for centering the explosion date prior distribution ) ; ` j ` a vector of integers specifying the filter i d of each photometric observation ; ` kcor_n ` , a matrix of pre - computed @xmath27-corrections for each filter , in magnitudes with spacing of 1 day ; ` fluxscale ` the zero - point of the luminosity unit system ( @xmath221 in the system we have employed ) ; and ` duringseason ` , a boolean value specifying whether the object exploded within or between observing seasons , for selection of the explosion date prior distribution parameters .
the calculation of the model light curve flux and application of the @xmath27-correction values is performed in the ` transformed parameters ` section , and the prior and likelihood calculations are performed in the ` model ` section .
certain vector - valued prior distribution parameters are specified in the ` transformed data ` section for convenience .
the _ stan _ model is then compiled and run @xcite to yield mcmc samples from the posterior distribution of light curve parameters .
we have used pystan version 2.2.0 .
we configured the _ no - u - turn sampler _ to use fixed @xmath222 initialization of the parameter values , an adaptation phase of 250 steps , a maximum treedepth of 22 , and otherwise employed the default sampler parameters . using this configuration ,
we achieve a gelman - rubin between - to - within chain variance ratio of @xmath223 for 95% of parameters , indicating excellent convergence .
the typical run time is 1 minute per chain .
.... data { int < lower=0 > n_obs ; int
< lower=0 > n_filt ; vector[n_obs ] t ; vector[n_obs ] fl ; vector[n_obs ] dfl ; real z ; real t0_mean ; int < lower=1,upper = n_filt > j[n_obs ] ; int < lower=0 > kcor_n ; real kcor[n_filt , kcor_n ] ; real < lower=0 > fluxscale ; real < lower=0,upper=1 > duringseason ; } transformed data { vector[n_filt ] prior_tp ; vector[n_filt ] prior_sig_tp ; vector[n_filt ] prior_lbeta1 ; vector[n_filt ] prior_sig_lbeta1 ; vector[n_filt ] prior_lbeta2 ; vector[n_filt ] prior_sig_lbeta2 ; prior_tp[1 ] < - log(5 ) ; prior_tp[2 ] < - log(8 ) ; prior_tp[3 ] < - log(14 ) ; prior_tp[4 ] < - log(20 ) ; prior_tp[5 ] < - log(30 ) ; prior_sig_tp[1 ] < - 0.3 ; prior_sig_tp[2 ] < - 0.3 ; prior_sig_tp[3 ] < - 0.3 ; prior_sig_tp[4 ] < - 0.3 ; prior_sig_tp[5 ] < - 0.3 ; prior_lbeta1[1 ] < - -2.1 ; prior_lbeta1[2 ] < - -2.3 ; prior_lbeta1[3 ] < - -3.3 ; prior_lbeta1[4 ] < - -3.8 ; prior_lbeta1[5 ] < - -4.0 ; prior_sig_lbeta1[1 ] < - 0.6 ; prior_sig_lbeta1[2 ] < - 0.8 ; prior_sig_lbeta1[3 ] < - 1.2 ; prior_sig_lbeta1[4 ] < - 1.5 ; prior_sig_lbeta1[5 ] < - 2 ; prior_lbeta2[1 ] < - -3.4 ; prior_lbeta2[2 ] < - -4 ; prior_lbeta2[3 ] < - -4.1 ; prior_lbeta2[4 ] < - -4.4 ; prior_lbeta2[5 ] < - -4.9 ; prior_sig_lbeta2[1 ] < - 1 ; prior_sig_lbeta2[2 ] < - 1.2 ; prior_sig_lbeta2[3 ] < - 1.2 ; prior_sig_lbeta2[4 ] < - 1.5 ; prior_sig_lbeta2[5 ] < - 1.5 ; } parameters { real pt0 ; vector < lower=0>[n_filt ] t1 ; vector < lower=0>[n_filt ] t2 ; vector < lower=0>[n_filt ] td ; vector < lower=0>[n_filt ] tp ; vector < upper=0>[n_filt ] lalpha ; vector < upper=0>[n_filt ] lbeta1 ; vector < upper=0>[n_filt ] lbeta2 ; vector < upper=0>[n_filt ] lbetadn ; vector < upper=0>[n_filt ] lbetadc ; vector < lower=0>[n_filt ] mp ; vector[n_filt ] yb ; vector < lower=0>[n_filt ] v ; } transformed parameters { vector[n_obs ] mm ; vector[n_obs ] dm ; vector < lower=0>[n_filt ] m1 ; vector < lower=0>[n_filt ] m2 ; vector < lower=0>[n_filt ] md ; m1 < - mp ./ exp ( exp(lbeta1 ) . * tp ) ; m2 < - mp . * exp ( -exp(lbeta2 ) . *
t2 ) ; md < - m2 . * exp ( -exp(lbetadn ) .
* td ) ; for ( n in 1:n_obs ) { real n_snc ; int kc_up ; int kc_down ; real t_exp ; int j ; int k ; real mm_1 ; real mm_2 ; real mm_3 ; real mm_4 ; real mm_5 ; real mm_6 ; j < - j[n ] ; t_exp < - ( t[n ] - ( t0_mean + pt0 ) ) / ( 1 + z ) ; if ( t_exp<0 ) { mm_1 < - yb[j ] ; } else { mm_1 < - 0 ; } if ( ( t_exp>=0 ) & & ( t_exp < t1[j ] ) ) { mm_2 < - yb[j ] + m1[j ] * pow(t_exp / t1[j ] , exp(lalpha[j ] ) ) ; } else { mm_2 < - 0 ; } if ( ( t_exp > = t1[j ] ) & & ( t_exp < t1[j ] + tp[j ] ) ) { mm_3 < - yb[j ] + m1[j ] * exp(exp(lbeta1[j ] ) * ( t_exp - t1[j ] ) ) ; } else { mm_3 < - 0 ; } if ( ( t_exp > = t1[j ] + tp[j ] ) & & ( t_exp < t1[j ] + tp[j ] + t2[j ] ) ) { mm_4 < - yb[j ] + mp[j ] * exp(-exp(lbeta2[j ] ) * ( t_exp - t1[j ] - tp[j ] ) ) ; } else { mm_4 < - 0 ; } if ( ( t_exp > = t1[j ] + tp[j ] + t2[j ] ) & & ( t_exp < t1[j ] + tp[j ] + t2[j ] + td[j ] ) ) { mm_5 < - yb[j ] + m2[j ] * exp(-exp(lbetadn[j ] ) * ( t_exp - t1[j ] - tp[j ] - t2[j ] ) ) ; } else { mm_5 < - 0 ; } if ( t_exp > = t1[j ] + tp[j ] + t2[j ] + td[j ] ) { mm_6 < - yb[j ] + md[j ] * exp(-exp(lbetadc[j ] ) * ( t_exp - t1[j ] - tp[j ] - t2[j ] - td[j ] ) ) ; } else { mm_6 < - 0 ; } dm[n ] < - sqrt(pow(dfl[n],2 ) + pow(v[j],2 ) ) ; if ( t_exp<0 ) { n_snc < - 0 ; } else if ( t_exp < kcor_n-2 ) { kc_down < - 0 ; while ( ( kc_down+1 ) < t_exp ) { kc_down < - kc_down + 1 ; } kc_up < - kc_down+1 ; n_snc < - kcor[j , kc_down+1 ] + ( t_exp - floor(t_exp ) ) * ( kcor[j , kc_up+1]-kcor[j , kc_down+1 ] ) ; } else { n_snc
< - kcor[j , kcor_n ] ; } mm[n ] < - ( mm_1+mm_2+mm_3+mm_4+mm_5+mm_6 ) / ( pow(10 , n_snc/(-2.5 ) ) ) ; } } model { if ( duringseason = = 1 ) { pt0 ~ skew_normal(-1 , 1 , -0.5 ) ; } else { pt0 ~ skew_normal(-30 , 20 , -1 ) ; } t1 ~ lognormal(log(1 ) , 0.3 ) ; tp ~ lognormal(prior_tp , prior_sig_tp ) ; t2 ~ lognormal(log(100 ) , 0.3 ) ; td ~ lognormal(log(10 ) , 0.5 ) ; lalpha ~ normal(-1 , 0.3 ) ; lbeta1 ~ normal(prior_lbeta1 , prior_sig_lbeta1 ) ; lbeta2 ~ normal(prior_lbeta2 , prior_sig_lbeta2 ) ; lbetadn ~ normal(-3 , 0.5 ) ; lbetadc ~ normal(-5 , 1 ) ; mp ~ lognormal(log(1 ) , 0.7 ) ; yb ~ normal(0 , 0.3 ) ; v ~ cauchy(0 , 0.01 ) ; fl ~ normal(mm , dm ) ; } ....
below we reproduce the full probabilistic model describing a linear relation between bivariate data with intrinsic scatter , which we apply to the sn iip decline rate
peak magnitude in section [ sec : res : pcor ] , specified in the _ stan _ modeling language and following @xcite .
the model takes the following data as input : ` n ` , the number of bivariate observations ; ` x ` and ` y ` , the mean observed values of the first and second covariates ; ` dx ` and ` dy ` , the uncertainty ( standard deviation ) of each observation projected along the axis of each covariate ; and ` dxy ` the covariance of each observation . in the ` transformed parameters ` section , the sampled angle of the linear relation from the @xmath224-axis ( `
theta ` ) and its perpendicular distance ( ` b_perp ` ) from the origin are used to calculate the perpendicular distance of each observation from the ridgeline of the sampled relation and its projected variance .
the prior and likelihood calculations are performed in the ` model ` section , accounting for both the projected observational variance and the intrinsic variance ( ` v ` ) .
the ` b_perp ` and ` theta ` parameters are transformed into traditional linear slope and offset parameters ( ` m , b ` ) in the ` generated quantities ` section for convenience .
transformed parameters { unit_vector[2 ] v ; vector[n ] delta ; matrix[n,2 ] z ; cov_matrix[2 ] s[n ] ; vector[n ] sigma_squared ; v[1 ] < - -sin(theta ) ; v[2 ] < - cos(theta ) ; for ( n in 1:n ) { z[n,1 ] < - x[n ] ; z[n,2 ] < - y[n ] ; s[n][1,1 ] < - pow(dx[n],2 ) ; s[n][2,2 ] < - pow(dy[n],2 ) ; s[n][1,2 ] < - dxy[n ] ; s[n][2,1 ] < - s[n][1,2 ] ; sigma_squared[n ] < - v ' * s[n ] * v ; delta[n ] < - v ' * z[n ] ' - b_perp ; } } lllllllllll ps110a & 02 & 03:29:48.735 & 29:04:29.91 & 55207 & 0.071 & 2 & 2 & 5 & 4 & 0 + ps110b & 03 & 08:37:02.587 & + 43:44:16.85 & 55207 & 0.260 & 4 & 4 & 5 & 0 & 0 + ps110q & 06 & 12:19:37.533 & + 46:01:56.03 & 55212 & 0.057 & 15 & 17 & 16 & 13 & 0 + ps110 t & 04 & 09:55:30.959 & + 01:29:01.18 & 55214 & 0.093 & 11 & 11 & 11 & 11 & 0 + ps110ae & 04 & 10:02:09.742 & + 01:14:00.93 & 55242 & 0.086 & 4 & 5 & 8 & 6 & 0 + ps110lv & 09 & 22:13:00.969 & + 00:46:24.49 & 55389 & 0.127 & 10 & 21 & 22 & 23 & 0 + ps110 mb & 08 & 16:11:44.435 & + 56:05:08.39 & 55389 & 0.064 & 2 & 12 & 13 & 10 & 0 + ps110zu & 01 & 02:18:54.489 & 04:14:42.96 & 55407 & 0.059 & 11 & 18 & 21 & 24 & 0 + ps110abd & 01 & 02:20:04.175 & 03:30:46.10 & 55414 & 0.132 & 5 & 15 & 18 & 16 & 0 + ps110agh & 01 & 02:21:04.532 & 03:06:14.81 & 55428 & 0.141 & 8 & 12 & 14 & 18 & 0 + ps110awc & 02 & 03:30:12.902 & 28:38:22.19 & 55467 & 0.123 & 8 & 11 & 12 & 15 & 0 + ps110biu & 09 & 22:14:20.250 & 00:19:00.10 & 55499 & 0.184 & 2 & 3 & 5 & 2 & 0 + ps110blk & 01 & 02:22:51.457 & 03:24:18.58 & 55510 & 0.074 & 7 & 6 & 9 & 5 & 0 + ps111u & 05 & 10:41:11.486 & + 58:14:11.61 & 55566 & 0.040 & 7 & 8 & 7 & 5 & 0 + ps111z & 04 & 10:03:16.250 & + 03:23:54.14 & 55566 & 0.105 & 8 & 9 & 11 & 7 & 0 + ps111ai & 04 & 09:56:42.562 & + 00:55:47.69 & 55567 & 0.092 & 3 & 6 & 7 & 5 & 0 + ps111ak & 02 & 03:33:29.222 & 28:45:26.54 & 55568 & 0.120 & 0 & 1 & 2 & 1 & 0 + ps111bo & 04 & 09:56:45.045 & + 00:58:08.29 & 55576 & 0.093 & 8 & 12 & 10 & 10 & 0 + ps111fl & 07 & 14:05:19.148 & + 53:07:24.25 & 55597 & 0.096 & 5 & 9 & 14 & 8 & 0 + ps111jg & 06 & 12:21:42.116 & + 47:23:00.01 & 55597 & 0.082 & 0 & 1 & 12 & 1 & 0 + ps111kf & 08 & 16:09:05.620 & + 54:07:19.71 & 55602 & 0.210 & 1 & 3 & 4 & 2 & 0 + ps111li & 06 & 12:16:14.114 & + 46:20:45.99 & 55605 & 0.068 & 4 & 11 & 0 & 9 & 0 + ps111lj & 06 & 12:16:56.601 & + 46:18:56.29 & 55605 & 0.070 & 12 & 14 & 16 & 14 & 0 + ps111lr & 03 &
08:37:58.889 & + 43:48:22.28 & 55615 & 0.055 & 4 & 4 & 7 & 5 & 0 + ps111lq & 04 & 09:57:11.372 & + 02:32:40.35 & 55615 & 0.101 & 8 & 9 & 7 & 5 & 0 + ps111md & 03 & 08:41:11.752 & + 43:22:51.45 & 55633 & 0.145 & 4 & 4 & 6 & 4 & 0 + ps111pf & 07 & 14:18:29.828 & + 53:43:31.79 & 55636 & 0.148 & 4 & 9 & 19 & 7 & 0 + ps111qe & 07 & 14:19:13.704 & + 54:09:15.03 & 55648 & 0.191 & 7 & 9 & 18 & 8 & 0 + ps111si & 05 & 10:43:40.661 & + 58:02:38.18 & 55666 & 0.133 & 0 & 2 & 5 & 2 & 0 + ps111sm & 07 & 14:20:18.612 & + 53:59:28.70 & 55666 & 0.209 & 4 & 4 & 9 & 2 & 0 + ps111ua & 08 & 16:12:00.684 & + 54:41:31.66 & 55675 & 0.236 & 2 & 3 & 7 & 1 & 0
+ ps111uv & 05 & 10:48:48.970 & + 57:47:35.99 & 55675 & 0.073 & 6 & 5 & 7 & 7 & 0 + ps111vn & 07 & 14:19:17.217 & + 53:45:07.20 & 55675 & 0.168 & 3 & 5 & 12 & 2 & 0 + ps111wj & 04 & 10:00:49.340 & + 02:52:46.68 & 55675 & 0.077 & 3 & 3 & 3 & 2 & 0 + ps111yh & 07 & 14:11:11.466 & + 51:59:12.63 & 55705 & 0.146 & 4 & 8 & 13 & 8 & 0 + ps111amh & 09 & 22:11:47.815 & + 01:09:00.95 & 55795 & 0.077 & 11 & 9 & 4 & 6 & 5 + ps111amv & 08 & 16:12:40.632 & + 56:11:18.38 & 55803 & 0.052 & 1 & 1 & 3 & 2 & 3 + ps111anm & 10 & 23:29:26.246 & 01:51:18.09 & 55775 & 0.061 & 3 & 20 & 29 & 20 & 5 + ps111aof & 10 & 23:25:59.575 & 00:30:55.16 & 55792 & 0.095 & 30 & 36 & 37 & 29 & 14 + ps111apd & 10 & 23:28:56.422 & 01:05:18.45 & 55800 & 0.043 & 25 & 26 & 23 & 30 & 13 + ps111asz & 02 & 03:34:46.879 & 29:12:32.51 & 55831 & 0.103 & 11 & 13 & 22 & 21 & 6 + ps111awd & 02 & 03:34:37.528 & 27:04:31.47 & 55852 & 0.068 & 9 & 3 & 14 & 11 & 13 + ps111azd & 10 & 23:31:47.673 & 01:25:51.41 & 55883 & 0.059 & 5 & 9 & 0 & 11 & 3 + ps111azi & 02 & 03:37:45.601 & 28:14:25.71 & 55883 & 0.180 & 4 & 4 & 9 & 11 & 7 + ps111azt & 01 & 02:27:34.018 & 03:50:34.07 & 55883 & 0.175 & 4 & 4 & 9 & 8 & 0 + ps111azv & 03 & 08:39:58.993 & + 45:11:36.90 & 55883 & 0.145 & 7 & 12 & 18 & 24 & 6 + ps111bbg & 03 & 08:49:51.815 & + 44:41:17.26 & 55893 & 0.132 & 4 & 15 & 17 & 24 & 8 + ps112c & 05 & 10:36:06.755 & + 57:47:01.85 & 55911 & 0.040 & 6 & 13 & 23 & 30 & 4 + ps112d & 05 & 10:39:55.485 & + 58:49:27.53 & 55911 & 0.123 & 11 & 16 & 18 & 30 & 9 + ps112y & 04 & 09:58:01.639 & + 00:43:14.16 & 55941 & 0.093 & 8 & 8 & 14 & 10 & 6 + ps112bt & 04 & 10:05:16.844 & + 01:40:39.57 & 55948 & 0.042 & 8 & 11 & 15 & 13 & 5 + ps112rk & 06 & 12:25:55.129 & + 47:36:06.38 & 56000 & 0.079 & 11 & 10 & 23 & 25 & 6 + ps112wn & 04 & 09:59:11.437 & + 01:08:38.27 & 56013 & 0.107 & 4 & 6 & 12 & 15 & 2 + ps112zd & 04 & 09:56:48.772 & + 01:01:45.56 & 56029 & 0.104 & 5 & 5 & 13 & 7 & 2 + ps112zp & 06 & 12:14:45.637 & + 48:19:13.49 & 56029 & 0.101 & 8 & 6 & 9 & 9 & 2 + ps112auw & 08 & 16:15:35.435 & + 56:11:26.92 & 56091 & 0.096 & 3 & 3 & 6 & 4 & 7 + ps112bku & 09 & 22:16:40.194 & 01:08:24.35 & 56166 & 0.096 & 7 & 11 & 15 & 11 & 8 + ps112bmp & 02 & 03:29:28.318 & 28:42:15.61 & 56207 & 0.070 & 10 & 12 & 18 & 23 & 21 + ps112bml & 02 & 03:27:04.863 & 27:17:15.73 & 56207 & 0.044 & 5 & 10 & 15 & 12 & 1 + ps112blz & 01 & 02:21:10.849 & 05:27:22.91 & 56210 & 0.086 & 9 & 12 & 16 & 24 & 13 + ps112boi & 09 & 22:22:40.280 & 00:07:21.02 & 56210 & 0.145 & 4 & 6 & 8 & 6 & 1 + ps112bza & 03 & 08:36:23.991 & + 43:38:26.14 & 56235 & 0.109 & 9 & 13 & 26 & 26 & 19 + ps112cey & 01 & 02:21:27.462 & 03:18:56.45 & 56240 & 0.140 & 3 & 5 & 9 & 13 & 4 + ps113hp & 04 & 10:04:16.974 & + 02:35:50.61 & 56326 & 0.149 & 11 & 12 & 19 & 20 & 5 + ps113vs & 04 & 09:57:24.212 & + 03:34:46.75 & 56353 & 0.079 & 6 & 8 & 17 & 17 & 6 + ps113wr & 07 & 14:15:57.954 & + 52:03:10.57 & 56385 & 0.077 & 10 & 16 & 27 & 29 & 27 + ps113alz & 04 & 09:57:52.631 & + 01:52:21.07 & 56399 & 0.098 & 1 & 1 & 7 & 5 & 5 + ps113arp & 06 & 12:18:25.108 & + 46:37:01.30 & 56408 & 0.167 & 6 & 9 & 14 & 12 & 0 + ps113ave & 07 & 14:08:52.904 & + 53:15:25.98 & 56417 & 0.185 & 13 & 12 & 21 & 24 & 16 + ps113atm & 04 & 09:59:16.605 & + 03:18:28.43 & 56417 & 0.240 & 1 & 1 & 5 & 5 & 1 + ps113bvb & 09 & 22:18:54.478 & + 01:41:32.70 & 56478 & 0.113 & 8 & 11 & 14 & 20 & 9 + ps113cqb & 10 & 23:23:35.356 & 00:20:33.23 & 56513 & 0.097 & 4 & 9 & 16 & 26 & 6 + ps113cua & 01 & 02:24:04.137 & 04:21:25.21 & 56564 & 0.095 & 13 & 16 & 17 & 20 & 7 + ps113eck & 04 & 09:54:25.733 & + 02:19:30.73 & 56624 & 0.121 & 3 & 4 & 8 & 11 & 9 + ps113esn & 04 & 09:57:23.048 & + 02:31:09.22 & 56659 & 0.130 & 3 & 3 & 7 & 3 & 0 lrrrrrrrrrrrr + ps110a & @xmath225 & @xmath226 & @xmath227 & @xmath228 & @xmath229 & @xmath230 & @xmath231 & @xmath232 & @xmath233 & @xmath234 & @xmath235 & @xmath236 + ps110b & @xmath237 & @xmath226 & @xmath238 & @xmath239 & @xmath240 & @xmath241 & @xmath242 & @xmath243 & @xmath244 & @xmath245 & @xmath246 & @xmath247 + ps110q & @xmath248 & @xmath226 & @xmath227 & @xmath249 & @xmath250 & @xmath251 & @xmath242 & @xmath252 & @xmath253 & @xmath254 & @xmath255 & @xmath256 + ps110 t & @xmath257 & @xmath226 & @xmath258 & @xmath259 & @xmath260 & @xmath261 & @xmath262 & @xmath232 & @xmath263 & @xmath234 & @xmath264 & @xmath265 + ps110ae & @xmath266 & @xmath267 & @xmath268 & @xmath269 & @xmath240 & @xmath270 & @xmath271 & @xmath272 & @xmath273 & @xmath274 & @xmath275 & @xmath236 + ps110lv & @xmath276 & @xmath226 & @xmath277 & @xmath249 & @xmath240 & @xmath278 & @xmath262 & @xmath279 & @xmath280 & @xmath245 & @xmath281 & @xmath282 + ps110 mb & @xmath283 & @xmath226 & @xmath227 & @xmath284 & @xmath240 & @xmath285 & @xmath262 & @xmath252 & @xmath286 & @xmath234 & @xmath287 & @xmath288 + ps110zu & @xmath289 & @xmath226 & @xmath277 & @xmath290 & @xmath291 & @xmath292 & @xmath242 & @xmath279 & @xmath293 & @xmath294 & @xmath295 & @xmath296 + ps110abd & @xmath297 & @xmath226 & @xmath277 & @xmath298 & @xmath240 & @xmath285 & @xmath242 & @xmath279 & @xmath299 & @xmath234 & @xmath300 & @xmath301 + ps110agh & @xmath302 & @xmath226 & @xmath303 & @xmath304 & @xmath240 & @xmath305 & @xmath242 & @xmath279 & @xmath306 & @xmath234 & @xmath307 & @xmath308 + ps110awc & @xmath309 & @xmath226 & @xmath310 & @xmath311 & @xmath312 & @xmath313 & @xmath231 & @xmath314 & @xmath315 & @xmath234 & @xmath316 & @xmath317 + ps110biu & @xmath318 & @xmath226 & @xmath319 & @xmath320 & @xmath240 & @xmath321 & @xmath322 & @xmath252 & @xmath323 & @xmath234 & @xmath324 & @xmath325 + ps110blk & @xmath326 & @xmath226 & @xmath327 & @xmath328 & @xmath240 & @xmath329 & @xmath231 & @xmath252 & @xmath330 & @xmath331 & @xmath332 & @xmath333 + ps111u & @xmath334 & @xmath226 & @xmath335 & @xmath336 & @xmath337 & @xmath338 & @xmath242 & @xmath279 & @xmath339 & @xmath340 & @xmath341 & @xmath342 + ps111z & @xmath343 & @xmath226 & @xmath277 & @xmath311 & @xmath240 & @xmath241 & @xmath242 & @xmath279 & @xmath344 & @xmath234 & @xmath345 & @xmath346 + ps111ai & @xmath347 & @xmath226 & @xmath277 & @xmath348 & @xmath240 & @xmath349 & @xmath242 & @xmath279 & @xmath350 & @xmath234 & @xmath351 & @xmath282 + ps111ak & @xmath352 & @xmath226 & @xmath277 & @xmath353 & @xmath240 & @xmath354 & @xmath262 & @xmath279 & @xmath355 & @xmath245 & @xmath356 & @xmath357 + ps111bo & @xmath358 & @xmath226 & @xmath277 & @xmath359 & @xmath240 & @xmath354 & @xmath242 & @xmath279 & @xmath360 & @xmath274 & @xmath361 & @xmath362 + ps111fl & @xmath363 & @xmath226 & @xmath335 & @xmath364 & @xmath229 & @xmath365 & @xmath242 & @xmath279 & @xmath366 & @xmath367 & @xmath368 & @xmath369 + ps111jg & @xmath370 & @xmath226 & @xmath371 & @xmath372 & @xmath240 & @xmath329 & @xmath242 & @xmath279 & @xmath373 & @xmath234 & @xmath374 & @xmath375 + ps111kf & @xmath376 & @xmath226 & @xmath377 & @xmath378 & @xmath229 & @xmath379 & @xmath242 & @xmath380 & @xmath381 & @xmath234 & @xmath382 & @xmath383 + ps111li & @xmath384 & @xmath226 & @xmath385 & @xmath259 & @xmath240 & @xmath386 & @xmath262 & @xmath252 & @xmath387 & @xmath234 & @xmath388 & @xmath389 + ps111lj & @xmath390 & @xmath226 & @xmath268 & @xmath359 & @xmath391 & @xmath392 & @xmath322 & @xmath252 & @xmath393 & @xmath394 & @xmath395 & @xmath362 + ps111lr & @xmath396 & @xmath226 & @xmath303 & @xmath397 & @xmath240 & @xmath386 & @xmath242 & @xmath279 & @xmath398 & @xmath245 & @xmath399 & @xmath400 + ps111lq & @xmath401 & @xmath226 & @xmath402 & @xmath403 & @xmath240 & @xmath404 & @xmath322 & @xmath252 & @xmath405 & @xmath245 & @xmath406 & @xmath407 + ps111md & @xmath408 & @xmath226 & @xmath277 & @xmath311 & @xmath240 & @xmath305 & @xmath262 & @xmath279 & @xmath409 & @xmath245 & @xmath410 & @xmath411 + ps111pf & @xmath412 & @xmath226 & @xmath413 & @xmath298 & @xmath240 & @xmath329 & @xmath262 & @xmath279 & @xmath414 & @xmath245 & @xmath415 & @xmath416 + ps111qe & @xmath417 & @xmath226 & @xmath418 & @xmath419 & @xmath420 & @xmath421 & @xmath322 & @xmath252 & @xmath422 & @xmath423 & @xmath424 & @xmath425 + ps111si & @xmath426 & @xmath226 & @xmath277 & @xmath427 & @xmath240 & @xmath329 & @xmath428 & @xmath279 & @xmath429 & @xmath245 & @xmath430 & @xmath431 + ps111sm & @xmath432 & @xmath226 & @xmath268 & @xmath433 & @xmath240 & @xmath434 & @xmath262 & @xmath252 & @xmath435 & @xmath245 & @xmath436 & @xmath437 + ps111ua & @xmath438 & @xmath226 & @xmath371 & @xmath439 & @xmath240 & @xmath313 & @xmath242 & @xmath279 & @xmath440 & @xmath234 & @xmath441 & @xmath442 + ps111uv & @xmath443 & @xmath226 & @xmath444 & @xmath259 & @xmath240 & @xmath354 & @xmath322 & @xmath445 & @xmath446 & @xmath234 & @xmath447 & @xmath400 + ps111vn & @xmath448 & @xmath226 & @xmath449 & @xmath450 & @xmath240 & @xmath329 & @xmath322 & @xmath252 & @xmath355 & @xmath245 & @xmath451 & @xmath452 + ps111wj & @xmath453 & @xmath226 & @xmath454 & @xmath455 & @xmath240 & @xmath386 & @xmath262 & @xmath252 & @xmath456 & @xmath234 & @xmath457 & @xmath458 + ps111yh & @xmath459 & @xmath226 & @xmath277 & @xmath311 & @xmath240 & @xmath386 & @xmath242 & @xmath279 & @xmath460 & @xmath245 & @xmath461 & @xmath462 + ps111amh & @xmath463 & @xmath226 & @xmath464 & @xmath259 & @xmath465 & @xmath466 & @xmath322 & @xmath445 & @xmath467 & @xmath468 & @xmath469 & @xmath470 + ps111amv & @xmath471 & @xmath472 & @xmath473 & @xmath474 & @xmath475 & @xmath386 & @xmath262 & @xmath380 & @xmath476 & @xmath245 & @xmath477 & @xmath478 + ps111anm & @xmath479 & @xmath226 & @xmath480 & @xmath481 & @xmath240 & @xmath386 & @xmath322 & @xmath252 & @xmath482 & @xmath234 & @xmath483 & @xmath484 + ps111aof & @xmath485 & @xmath226 & @xmath268 & @xmath486 & @xmath240 & @xmath251 & @xmath322 & @xmath445 & @xmath286 & @xmath245 & @xmath487 & @xmath488 + ps111apd & @xmath489 & @xmath226 & @xmath319 & @xmath490 & @xmath491 & @xmath492 & @xmath322 & @xmath252 & @xmath493 & @xmath494 & @xmath495 & @xmath496 + ps111asz & @xmath497 & @xmath226 & @xmath472 & @xmath498 & @xmath240 & @xmath329 & @xmath231 & @xmath252 & @xmath499 & @xmath234 & @xmath500 & @xmath501 + ps111awd & @xmath502 & @xmath226 & @xmath503 & @xmath397 & @xmath240 & @xmath504 & @xmath322 & @xmath445 & @xmath505 & @xmath234 & @xmath506 & @xmath507 + ps111azd & @xmath508 & @xmath226 & @xmath509 & @xmath510 & @xmath229 & @xmath511 & @xmath322 & @xmath252 & @xmath512 & @xmath513 & @xmath514 & @xmath515 + ps111azi & @xmath516 & @xmath226 & @xmath277 & @xmath320 & @xmath240 & @xmath517 & @xmath262 & @xmath279 & @xmath518 & @xmath245 & @xmath519 & @xmath520 + ps111azt & @xmath521 & @xmath226 & @xmath480 & @xmath522 & @xmath523 & @xmath524 & @xmath428 & @xmath232 & @xmath525 & @xmath245 & @xmath526 & @xmath527 + ps111azv & @xmath528 & @xmath226 & @xmath335 & @xmath359 & @xmath240 & @xmath529 & @xmath231 & @xmath279 & @xmath530 & @xmath245 & @xmath531 & @xmath317 + ps111bbg & @xmath532 & @xmath226 & @xmath277 & @xmath533 & @xmath240 & @xmath534 & @xmath242 & @xmath232 & @xmath535 & @xmath234 & @xmath536 & @xmath537 + ps112c & @xmath538 & @xmath226 & @xmath277 & @xmath539 & @xmath420 & @xmath540 & @xmath262 & @xmath279 & @xmath541 & @xmath542 & @xmath543 & @xmath544 + ps112d & @xmath545 & @xmath226 & @xmath277 & @xmath546 & @xmath240 & @xmath547 & @xmath262 & @xmath232 & @xmath548 & @xmath245 & @xmath549 & @xmath550 + ps112y & @xmath551 & @xmath226 & @xmath277 & @xmath552 & @xmath553 & @xmath554 & @xmath242 & @xmath279 & @xmath555 & @xmath556 & @xmath557 & @xmath558 + ps112bt & @xmath559 & @xmath226 & @xmath277 & @xmath560 & @xmath475 & @xmath561 & @xmath242 & @xmath232 & @xmath562 & @xmath563 & @xmath564 & @xmath342 + ps112rk & @xmath565 & @xmath226 & @xmath566 & @xmath567 & @xmath312 & @xmath568 & @xmath322 & @xmath445 & @xmath569 & @xmath245 & @xmath570 & @xmath571 + ps112wn & @xmath572 & @xmath226 & @xmath573 & @xmath574 & @xmath240 & @xmath321 & @xmath242 & @xmath314 & @xmath575 & @xmath513 & @xmath576 & @xmath407 + ps112zd & @xmath577 & @xmath226 & @xmath578 & @xmath419 & @xmath240 & @xmath547 & @xmath262 & @xmath279 & @xmath579 & @xmath245 & @xmath580 & @xmath581 + ps112zp & @xmath582 & @xmath226 & @xmath335 & @xmath419 & @xmath523 & @xmath261 & @xmath242 & @xmath468 & @xmath583 & @xmath245 & @xmath584 & @xmath585 + ps112auw & @xmath586 & @xmath226 & @xmath277 & @xmath378 & @xmath240 & @xmath587 & @xmath262 & @xmath232 & @xmath588 & @xmath245 & @xmath589 & @xmath590 + ps112bku & @xmath591 & @xmath592 & @xmath593 & @xmath594 & @xmath240 & @xmath595 & @xmath231 & @xmath232 & @xmath596 & @xmath597 & @xmath598 & @xmath599 + ps112bmp & @xmath600 & @xmath226 & @xmath277 & @xmath601 & @xmath602 & @xmath603 & @xmath322 & @xmath252 & @xmath604 & @xmath367 & @xmath605 & @xmath606 + ps112bml & @xmath607 & @xmath226 & @xmath277 & @xmath402 & @xmath240 & @xmath313 & @xmath262 & @xmath608 & @xmath609 & @xmath234 & @xmath610 & @xmath611 + ps112blz & @xmath612 & @xmath226 & @xmath227 & @xmath613 & @xmath614 & @xmath615 & @xmath231 & @xmath252 & @xmath616 & @xmath245 & @xmath617 & @xmath618 + ps112boi & @xmath619 & @xmath226 & @xmath566 & @xmath620 & @xmath291 & @xmath621 & @xmath271 & @xmath445 & @xmath622 & @xmath623 & @xmath624 & @xmath317 + ps112bza & @xmath625 & @xmath226 & @xmath258 & @xmath626 & @xmath614 & @xmath627 & @xmath231 & @xmath279 & @xmath628 & @xmath245 & @xmath629 & @xmath630 + ps112cey & @xmath631 & @xmath226 & @xmath632 & @xmath433 & @xmath240 & @xmath404 & @xmath231 & @xmath608 & @xmath633 & @xmath234 & @xmath634 & @xmath411 + ps113hp & @xmath635 & @xmath226 & @xmath636 & @xmath637 & @xmath240 & @xmath251 & @xmath231 & @xmath314 & @xmath638 & @xmath245 & @xmath639 & @xmath640 + ps113vs & @xmath641 & @xmath226 & @xmath480 & @xmath552 & @xmath229 & @xmath642 & @xmath262 & @xmath279 & @xmath643 & @xmath234 & @xmath644 & @xmath645 + ps113wr & @xmath646 & @xmath226 & @xmath303 & @xmath647 & @xmath648 & @xmath649 & @xmath242 & @xmath279 & @xmath650 & @xmath651 & @xmath652 & @xmath653 + ps113alz & @xmath654 & @xmath655 & @xmath371 & @xmath656 & @xmath240 & @xmath657 & @xmath231 & @xmath658 & @xmath659 & @xmath234 & @xmath660 & @xmath236 + ps113arp & @xmath661 & @xmath226 & @xmath291 & @xmath533 & @xmath240 & @xmath662 & @xmath322 & @xmath252 & @xmath663 & @xmath245 & @xmath664 & @xmath665 + ps113ave & @xmath666 & @xmath226 & @xmath303 & @xmath490 & @xmath667 & @xmath313 & @xmath262 & @xmath279 & @xmath668 & @xmath669 & @xmath670 & @xmath671 + ps113atm & @xmath672 & @xmath226 & @xmath673 & @xmath674 & @xmath240 & @xmath675 & @xmath242 & @xmath676 & @xmath677 & @xmath678 & @xmath679 & @xmath680 + ps113bvb & @xmath681 & @xmath226 & @xmath277 & @xmath269 & @xmath240 & @xmath682 & @xmath262 & @xmath279 & @xmath683 & @xmath234 & @xmath684 & @xmath685 + ps113cqb & @xmath686 & @xmath226 & @xmath566 & @xmath687 & @xmath240 & @xmath241 & @xmath262 & @xmath252 & @xmath688 & @xmath234 & @xmath576 & @xmath689 + ps113cua & @xmath690 & @xmath267 & @xmath509 & @xmath397 & @xmath312 & @xmath329 & @xmath322 & @xmath252 & @xmath691 & @xmath245 & @xmath692 & @xmath693 + ps113eck & @xmath694 & @xmath226 & @xmath480 & @xmath498 & @xmath240 & @xmath305 & @xmath242 & @xmath279 & @xmath695 & @xmath245 & @xmath696 & @xmath697 + ps113esn & @xmath698 & @xmath226 & @xmath277 & @xmath699 & @xmath240 & @xmath329 & @xmath242 & @xmath279 & @xmath700 & @xmath245 & @xmath701 & @xmath702 + + ps110a & @xmath225 & @xmath226 & @xmath703 & @xmath704 & @xmath250 & @xmath705 & @xmath262 & @xmath706 & @xmath707 & @xmath367 & @xmath708 & @xmath689 + ps110b & @xmath237 & @xmath226 & @xmath648 & @xmath419 & @xmath523 & @xmath329 & @xmath231 & @xmath243 & @xmath709 & @xmath234 & @xmath710 & @xmath711 + ps110q & @xmath248 & @xmath226 & @xmath712 & @xmath713 & @xmath714 & @xmath715 & @xmath262 & @xmath380 & @xmath716 & @xmath423 & @xmath717 & @xmath718 + ps110 t & @xmath257 & @xmath226 & @xmath719 & @xmath720 & @xmath721 & @xmath722 & @xmath723 & @xmath232 & @xmath724 & @xmath725 & @xmath726 & @xmath727 + ps110ae & @xmath266 & @xmath226 & @xmath728 & @xmath397 & @xmath523 & @xmath241 & @xmath322 & @xmath279 & @xmath729 & @xmath730 & @xmath731 & @xmath732 + ps110lv & @xmath276 & @xmath226 & @xmath733 & @xmath734 & @xmath291 & @xmath735 & @xmath242 & @xmath736 & @xmath737 & @xmath623 & @xmath738 & @xmath739 + ps110 mb & @xmath283 & @xmath226 & @xmath740 & @xmath741 & @xmath229 & @xmath742 & @xmath428 & @xmath743 & @xmath744 & @xmath367 & @xmath745 & @xmath746 + ps110zu & @xmath289 & @xmath226 & @xmath747 & @xmath748 & @xmath749 & @xmath750 & @xmath242 & @xmath751 & @xmath752 & @xmath753 & @xmath754 & @xmath755 + ps110abd & @xmath297 & @xmath226 & @xmath756 & @xmath359 & @xmath757 & @xmath758 & @xmath231 & @xmath751 & @xmath759 & @xmath254 & @xmath760 & @xmath761 + ps110agh & @xmath302 & @xmath226 & @xmath762 & @xmath763 & @xmath327 & @xmath764 & @xmath262 & @xmath706 & @xmath765 & @xmath766 & @xmath767 & @xmath768 + ps110awc & @xmath309 & @xmath226 & @xmath769 & @xmath770 & @xmath771 & @xmath772 & @xmath262 & @xmath773 & @xmath774 & @xmath775 & @xmath776 & @xmath777 + ps110biu & @xmath318 & @xmath226 & @xmath778 & @xmath779 & @xmath240 & @xmath780 & @xmath322 & @xmath468 & @xmath781 & @xmath730 & @xmath782 & @xmath325 + ps110blk & @xmath326 & @xmath226 & @xmath511 & @xmath783 & @xmath784 & @xmath305 & @xmath271 & @xmath380 & @xmath785 & @xmath730 & @xmath786 & @xmath787 + ps111u & @xmath334 & @xmath226 & @xmath756 & @xmath713 & @xmath419 & @xmath788 & @xmath262 & @xmath706 & @xmath789 & @xmath790 & @xmath791 & @xmath342 + ps111z & @xmath343 & @xmath226 & @xmath792 & @xmath783 & @xmath769 & @xmath793 & @xmath262 & @xmath794 & @xmath795 & @xmath796 & @xmath797 & @xmath798 + ps111ai & @xmath347 & @xmath226 & @xmath799 & @xmath800 & @xmath801 & @xmath802 & @xmath231 & @xmath751 & @xmath803 & @xmath804 & @xmath805 & @xmath806 + ps111ak & @xmath352 & @xmath226 & @xmath807 & @xmath808 & @xmath240 & @xmath329 & @xmath428 & @xmath751 & @xmath809 & @xmath245 & @xmath810 & @xmath811 + ps111bo & @xmath358 & @xmath226 & @xmath812 & @xmath813 & @xmath491 & @xmath524 & @xmath242 & @xmath751 & @xmath814 & @xmath815 & @xmath816 & @xmath817 + ps111fl & @xmath363 & @xmath226 & @xmath747 & @xmath783 & @xmath818 & @xmath819 & @xmath428 & @xmath751 & @xmath820 & @xmath821 & @xmath822 & @xmath823 + ps111jg & @xmath370 & @xmath226 & @xmath812 & @xmath824 & @xmath312 & @xmath825 & @xmath242 & @xmath751 & @xmath826 & @xmath815 & @xmath827 & @xmath693 + ps111kf & @xmath376 & @xmath226 & @xmath756 & @xmath828 & @xmath721 & @xmath829 & @xmath242 & @xmath751 & @xmath830 & @xmath623 & @xmath831 & @xmath247 + ps111li & @xmath384 & @xmath226 & @xmath832 & @xmath833 & @xmath636 & @xmath386 & @xmath231 & @xmath706 & @xmath834 & @xmath245 & @xmath835 & @xmath836 + ps111lj & @xmath390 & @xmath226 & @xmath837 & @xmath838 & @xmath290 & @xmath403 & @xmath242 & @xmath706 & @xmath839 & @xmath840 & @xmath841 & @xmath842 + ps111lr & @xmath396 & @xmath226 & @xmath799 & @xmath843 & @xmath240 & @xmath386 & @xmath428 & @xmath751 & @xmath844 & @xmath234 & @xmath845 & @xmath846 + ps111lq & @xmath401 & @xmath226 & @xmath799 & @xmath713 & @xmath229 & @xmath354 & @xmath262 & @xmath751 & @xmath847 & @xmath245 & @xmath848 & @xmath849 + ps111md & @xmath408 & @xmath226 & @xmath850 & @xmath851 & @xmath240 & @xmath329 & @xmath262 & @xmath751 & @xmath435 & @xmath234 & @xmath852 & @xmath317 + ps111pf & @xmath412 & @xmath226 & @xmath747 & @xmath359 & @xmath853 & @xmath854 & @xmath231 & @xmath706 & @xmath855 & @xmath669 & @xmath856 & @xmath857 + ps111qe & @xmath417 & @xmath226 & @xmath858 & @xmath269 & @xmath859 & @xmath860 & @xmath322 & @xmath232 & @xmath861 & @xmath730 & @xmath862 & @xmath863 + ps111si & @xmath426 & @xmath226 & @xmath864 & @xmath865 & @xmath240 & @xmath313 & @xmath242 & @xmath866 & @xmath499 & @xmath245 & @xmath867 & @xmath868 + ps111sm & @xmath432 & @xmath226 & @xmath869 & @xmath870 & @xmath229 & @xmath386 & @xmath322 & @xmath232 & @xmath871 & @xmath245 & @xmath872 & @xmath873 + ps111ua & @xmath438 & @xmath226 & @xmath762 & @xmath269 & @xmath229 & @xmath251 & @xmath428 & @xmath874 & @xmath875 & @xmath367 & @xmath876 & @xmath877 + ps111uv & @xmath443 & @xmath226 & @xmath878 & @xmath879 & @xmath240 & @xmath880 & @xmath322 & @xmath314 & @xmath881 & @xmath234 & @xmath882 & @xmath846 + ps111vn & @xmath448 & @xmath226 & @xmath883 & @xmath884 & @xmath240 & @xmath329 & @xmath322 & @xmath676 & @xmath885 & @xmath234 & @xmath886 & @xmath887 + ps111wj & @xmath453 & @xmath226 & @xmath888 & @xmath889 & @xmath240 & @xmath890 & @xmath322 & @xmath252 & @xmath891 & @xmath245 & @xmath892 & @xmath893 + ps111yh & @xmath459 & @xmath226 & @xmath799 & @xmath713 & @xmath475 & @xmath261 & @xmath242 & @xmath751 & @xmath894 & @xmath234 & @xmath895 & @xmath896 + ps111amh & @xmath463 & @xmath226 & @xmath897 & @xmath898 & @xmath899 & @xmath900 & @xmath901 & @xmath445 & @xmath902 & @xmath234 & @xmath903 & @xmath904 + ps111amv & @xmath471 & @xmath226 & @xmath905 & @xmath906 & @xmath240 & @xmath329 & @xmath242 & @xmath874 & @xmath907 & @xmath234 & @xmath908 & @xmath909 + ps111anm & @xmath479 & @xmath226 & @xmath503 & @xmath910 & @xmath229 & @xmath911 & @xmath428 & @xmath912 & @xmath913 & @xmath367 & @xmath914 & @xmath515 + ps111aof & @xmath485 & @xmath226 & @xmath915 & @xmath269 & @xmath240 & @xmath587 & @xmath322 & @xmath232 & @xmath916 & @xmath274 & @xmath917 & @xmath918 + ps111apd & @xmath489 & @xmath226 & @xmath919 & @xmath920 & @xmath921 & @xmath922 & @xmath262 & @xmath314 & @xmath923 & @xmath924 & @xmath925 & @xmath926 + ps111asz & @xmath497 & @xmath226 & @xmath927 & @xmath928 & @xmath632 & @xmath929 & @xmath322 & @xmath930 & @xmath541 & @xmath931 & @xmath932 & @xmath933 + ps111awd & @xmath502 & @xmath226 & @xmath832 & @xmath800 & @xmath934 & @xmath935 & @xmath428 & @xmath751 & @xmath936 & @xmath623 & @xmath937 & @xmath938 + ps111azd & @xmath508 & @xmath226 & @xmath939 & @xmath712 & @xmath229 & @xmath940 & @xmath231 & @xmath941 & @xmath659 & @xmath367 & @xmath942 & @xmath943 + ps111azi & @xmath516 & @xmath226 & @xmath799 & @xmath259 & @xmath229 & @xmath642 & @xmath231 & @xmath751 & @xmath944 & @xmath234 & @xmath945 & @xmath282 + ps111azt & @xmath521 & @xmath226 & @xmath919 & @xmath946 & @xmath229 & @xmath947 & @xmath231 & @xmath314 & @xmath948 & @xmath949 & @xmath950 & @xmath317 + ps111azv & @xmath528 & @xmath226 & @xmath951 & @xmath952 & @xmath953 & @xmath954 & @xmath262 & @xmath706 & @xmath955 & @xmath956 & @xmath957 & @xmath958 + ps111bbg & @xmath532 & @xmath592 & @xmath959 & @xmath290 & @xmath312 & @xmath540 & @xmath960 & @xmath941 & @xmath961 & @xmath245 & @xmath962 & @xmath963 + ps112c & @xmath538 & @xmath226 & @xmath747 & @xmath647 & @xmath964 & @xmath965 & @xmath262 & @xmath751 & @xmath966 & @xmath967 & @xmath968 & @xmath969 + ps112d & @xmath545 & @xmath226 & @xmath919 & @xmath970 & @xmath327 & @xmath971 & @xmath262 & @xmath972 & @xmath973 & @xmath974 & @xmath975 & @xmath976 + ps112y & @xmath551 & @xmath226 & @xmath799 & @xmath977 & @xmath978 & @xmath979 & @xmath428 & @xmath751 & @xmath980 & @xmath981 & @xmath982 & @xmath265 + ps112bt & @xmath559 & @xmath226 & @xmath762 & @xmath983 & @xmath240 & @xmath984 & @xmath242 & @xmath751 & @xmath985 & @xmath245 & @xmath986 & @xmath987 + ps112rk & @xmath565 & @xmath226 & @xmath988 & @xmath989 & @xmath240 & @xmath990 & @xmath231 & @xmath380 & @xmath575 & @xmath513 & @xmath991 & @xmath992 + ps112wn & @xmath572 & @xmath226 & @xmath993 & @xmath994 & @xmath523 & @xmath995 & @xmath231 & @xmath380 & @xmath996 & @xmath234 & @xmath997 & @xmath998 + ps112zd & @xmath577 & @xmath226 & @xmath799 & @xmath999 & @xmath523 & @xmath1000 & @xmath231 & @xmath706 & @xmath1001 & @xmath234 & @xmath1002 & @xmath1003 + ps112zp & @xmath582 & @xmath226 & @xmath1004 & @xmath1005 & @xmath899 & @xmath1006 & @xmath262 & @xmath1007 & @xmath1008 & @xmath254 & @xmath1009 & @xmath1010 + ps112auw & @xmath586 & @xmath226 & @xmath1011 & @xmath1012 & @xmath240 & @xmath386 & @xmath231 & @xmath314 & @xmath1013 & @xmath245 & @xmath1014 & @xmath1015 + ps112bku & @xmath591 & @xmath1016 & @xmath1017 & @xmath770 & @xmath238 & @xmath1018 & @xmath1019 & @xmath1020 & @xmath1021 & @xmath730 & @xmath1022 & @xmath1023 + ps112bmp & @xmath600 & @xmath226 & @xmath1024 & @xmath1025 & @xmath310 & @xmath1026 & @xmath322 & @xmath252 & @xmath1027 & @xmath1028 & @xmath1029 & @xmath693 + ps112bml & @xmath607 & @xmath226 & @xmath1004 & @xmath574 & @xmath240 & @xmath1030 & @xmath231 & @xmath1031 & @xmath1032 & @xmath234 & @xmath1033 & @xmath1034 + ps112blz & @xmath612 & @xmath226 & @xmath778 & @xmath594 & @xmath1035 & @xmath1036 & @xmath322 & @xmath232 & @xmath1037 & @xmath234 & @xmath1038 & @xmath1039 + ps112boi & @xmath619 & @xmath226 & @xmath1040 & @xmath522 & @xmath240 & @xmath860 & @xmath231 & @xmath314 & @xmath659 & @xmath245 & @xmath1041 & @xmath1042 + ps112bza & @xmath625 & @xmath226 & @xmath850 & @xmath601 & @xmath1043 & @xmath1044 & @xmath242 & @xmath1045 & @xmath1046 & @xmath1047 & @xmath1048 & @xmath1049 + ps112cey & @xmath631 & @xmath226 & @xmath1050 & @xmath999 & @xmath240 & @xmath1051 & @xmath322 & @xmath232 & @xmath1052 & @xmath245 & @xmath1053 & @xmath849 + ps113hp & @xmath635 & @xmath226 & @xmath687 & @xmath713 & @xmath229 & @xmath1054 & @xmath262 & @xmath1055 & @xmath1056 & @xmath949 & @xmath1057 & @xmath1058 + ps113vs & @xmath641 & @xmath226 & @xmath850 & @xmath1059 & @xmath291 & @xmath1060 & @xmath231 & @xmath706 & @xmath1061 & @xmath1062 & @xmath1063 & @xmath1064 + ps113wr & @xmath646 & @xmath226 & @xmath1065 & @xmath561 & @xmath239 & @xmath1066 & @xmath242 & @xmath751 & @xmath1067 & @xmath1068 & @xmath1069 & @xmath389 + ps113alz & @xmath654 & @xmath226 & @xmath799 & @xmath1070 & @xmath240 & @xmath1071 & @xmath262 & @xmath1072 & @xmath1073 & @xmath513 & @xmath1074 & @xmath1075 + ps113arp & @xmath661 & @xmath226 & @xmath1076 & @xmath1077 & @xmath240 & @xmath329 & @xmath262 & @xmath706 & @xmath1078 & @xmath234 & @xmath1079 & @xmath1080 + ps113ave & @xmath666 & @xmath226 & @xmath1081 & @xmath1082 & @xmath240 & @xmath285 & @xmath242 & @xmath1055 & @xmath663 & @xmath245 & @xmath1083 & @xmath1084 + ps113atm & @xmath672 & @xmath226 & @xmath1085 & @xmath1086 & @xmath636 & @xmath682 & @xmath242 & @xmath751 & @xmath583 & @xmath245 & @xmath1087 & @xmath1088 + ps113bvb & @xmath681 & @xmath226 & @xmath1089 & @xmath928 & @xmath1090 & @xmath1091 & @xmath242 & @xmath706 & @xmath1092 & @xmath1093 & @xmath1094 & @xmath1095 + ps113cqb & @xmath686 & @xmath226 & @xmath1096 & @xmath720 & @xmath240 & @xmath1097 & @xmath1098 & @xmath468 & @xmath1099 & @xmath234 & @xmath1100 & @xmath1101 + ps113cua & @xmath690 & @xmath226 & @xmath1050 & @xmath1102 & @xmath1103 & @xmath365 & @xmath262 & @xmath743 & @xmath1104 & @xmath651 & @xmath1105 & @xmath1049 + ps113eck & @xmath694 & @xmath226 & @xmath1106 & @xmath1107 & @xmath1108 & @xmath1109 & @xmath242 & @xmath972 & @xmath1110 & @xmath1111 & @xmath1112 & @xmath1113 + ps113esn & @xmath698 & @xmath226 & @xmath1114 & @xmath1115 & @xmath240 & @xmath305 & @xmath242 & @xmath751 & @xmath881 & @xmath245 & @xmath1116 & @xmath1117 + + ps110a & @xmath225 & @xmath226 & @xmath1118 & @xmath1005 & @xmath312 & @xmath1119 & @xmath322 & @xmath243 & @xmath1120 & @xmath678 & @xmath1121 & @xmath400 + ps110b & @xmath237 & @xmath226 & @xmath1122 & @xmath824 & @xmath240 & @xmath354 & @xmath231 & @xmath866 & @xmath456 & @xmath245 & @xmath1123 & @xmath1124 + ps110q & @xmath248 & @xmath226 & @xmath1125 & @xmath928 & @xmath509 & @xmath1126 & @xmath231 & @xmath1127 & @xmath1128 & @xmath967 & @xmath1129 & @xmath1130 + ps110 t & @xmath257 & @xmath226 & @xmath1131 & @xmath1132 & @xmath636 & @xmath890 & @xmath231 & @xmath1133 & @xmath1134 & @xmath245 & @xmath1135 & @xmath1136 + ps110ae & @xmath266 & @xmath1137 & @xmath1138 & @xmath1139 & @xmath859 & @xmath1140 & @xmath1141 & @xmath1142 & @xmath1143 & @xmath949 & @xmath1144 & @xmath1145 + ps110lv & @xmath276 & @xmath226 & @xmath1146 & @xmath928 & @xmath1147 & @xmath1148 & @xmath242 & @xmath1149 & @xmath1150 & @xmath730 & @xmath1151 & @xmath1152 + ps110 mb & @xmath283 & @xmath226 & @xmath1153 & @xmath1154 & @xmath229 & @xmath1155 & @xmath242 & @xmath1156 & @xmath323 & @xmath730 & @xmath1157 & @xmath1158 + ps110zu & @xmath289 & @xmath226 & @xmath1159 & @xmath1160 & @xmath1161 & @xmath1162 & @xmath262 & @xmath1163 & @xmath1164 & @xmath1165 & @xmath1166 & @xmath1167 + ps110abd & @xmath297 & @xmath226 & @xmath1168 & @xmath1169 & @xmath268 & @xmath1170 & @xmath242 & @xmath1149 & @xmath1171 & @xmath623 & @xmath1172 & @xmath1173 + ps110agh & @xmath302 & @xmath226 & @xmath1174 & @xmath1175 & @xmath420 & @xmath251 & @xmath723 & @xmath1055 & @xmath1176 & @xmath1177 & @xmath1178 & @xmath798 + ps110awc & @xmath309 & @xmath226 & @xmath1179 & @xmath928 & @xmath391 & @xmath1180 & @xmath262 & @xmath1181 & @xmath1182 & @xmath1183 & @xmath1184 & @xmath1185 + ps110biu & @xmath318 & @xmath226 & @xmath1186 & @xmath552 & @xmath229 & @xmath1187 & @xmath322 & @xmath243 & @xmath1188 & @xmath367 & @xmath1189 & @xmath1190 + ps110blk & @xmath326 & @xmath226 & @xmath1191 & @xmath1192 & @xmath988 & @xmath1155 & @xmath242 & @xmath1193 & @xmath1194 & @xmath234 & @xmath1195 & @xmath1196 + ps111u & @xmath334 & @xmath226 & @xmath1159 & @xmath1197 & @xmath1198 & @xmath1199 & @xmath242 & @xmath1200 & @xmath1201 & @xmath1202 & @xmath1203 & @xmath342 + ps111z & @xmath343 & @xmath226 & @xmath1204 & @xmath1205 & @xmath420 & @xmath1206 & @xmath262 & @xmath1156 & @xmath1207 & @xmath1208 & @xmath1209 & @xmath1210 + ps111ai & @xmath347 & @xmath226 & @xmath1211 & @xmath1212 & @xmath801 & @xmath1213 & @xmath262 & @xmath1214 & @xmath1215 & @xmath1216 & @xmath1217 & @xmath1218 + ps111ak & @xmath352 & @xmath226 & @xmath1219 & @xmath1220 & @xmath240 & @xmath329 & @xmath262 & @xmath1149 & @xmath1221 & @xmath245 & @xmath1222 & @xmath1223 + ps111bo & @xmath358 & @xmath226 & @xmath228 & @xmath1224 & @xmath240 & @xmath329 & @xmath231 & @xmath1225 & @xmath1226 & @xmath234 & @xmath1227 & @xmath1228 + ps111fl & @xmath363 & @xmath226 & @xmath1229 & @xmath1230 & @xmath1231 & @xmath1232 & @xmath242 & @xmath1214 & @xmath1233 & @xmath766 & @xmath1234 & @xmath1235 + ps111jg & @xmath370 & @xmath226 & @xmath1236 & @xmath1237 & @xmath648 & @xmath838 & @xmath242 & @xmath1238 & @xmath1239 & @xmath678 & @xmath1240 & @xmath693 + ps111kf & @xmath376 & @xmath1241 & @xmath1168 & @xmath1242 & @xmath1243 & @xmath1244 & @xmath242 & @xmath1245 & @xmath1246 & @xmath725 & @xmath1247 & @xmath1248 + ps111li & @xmath384 & @xmath226 & @xmath1249 & @xmath1250 & @xmath240 & @xmath251 & @xmath242 & @xmath1149 & @xmath844 & @xmath234 & @xmath1251 & @xmath1252 + ps111lj & @xmath390 & @xmath226 & @xmath1253 & @xmath1254 & @xmath699 & @xmath534 & @xmath242 & @xmath1255 & @xmath1256 & @xmath1257 & @xmath1258 & @xmath1259 + ps111lr & @xmath396 & @xmath226 & @xmath1260 & @xmath1261 & @xmath636 & @xmath329 & @xmath242 & @xmath1156 & @xmath1262 & @xmath234 & @xmath1263 & @xmath1264 + ps111lq & @xmath401 & @xmath226 & @xmath1265 & @xmath770 & @xmath240 & @xmath354 & @xmath231 & @xmath1133 & @xmath1266 & @xmath245 & @xmath1267 & @xmath1268 + ps111md & @xmath408 & @xmath226 & @xmath1269 & @xmath977 & @xmath240 & @xmath285 & @xmath242 & @xmath1270 & @xmath1134 & @xmath245 & @xmath1271 & @xmath1190 + ps111pf & @xmath412 & @xmath226 & @xmath1272 & @xmath928 & @xmath240 & @xmath329 & @xmath960 & @xmath1273 & @xmath1274 & @xmath234 & @xmath1275 & @xmath1276 + ps111qe & @xmath417 & @xmath226 & @xmath1277 & @xmath311 & @xmath1278 & @xmath1279 & @xmath322 & @xmath1280 & @xmath1281 & @xmath1282 & @xmath1283 & @xmath1284 + ps111si & @xmath426 & @xmath226 & @xmath614 & @xmath1285 & @xmath240 & @xmath329 & @xmath262 & @xmath1286 & @xmath499 & @xmath245 & @xmath1287 & @xmath1288 + ps111sm & @xmath432 & @xmath226 & @xmath742 & @xmath1289 & @xmath240 & @xmath329 & @xmath322 & @xmath874 & @xmath1290 & @xmath245 & @xmath1291 & @xmath1292 + ps111ua & @xmath438 & @xmath226 & @xmath1168 & @xmath1293 & @xmath1294 & @xmath1295 & @xmath428 & @xmath1238 & @xmath1296 & @xmath1297 & @xmath1298 & @xmath1299 + ps111uv & @xmath443 & @xmath226 & @xmath1300 & @xmath1301 & @xmath240 & @xmath386 & @xmath960 & @xmath874 & @xmath1302 & @xmath234 & @xmath1303 & @xmath236 + ps111vn & @xmath448 & @xmath226 & @xmath1153 & @xmath359 & @xmath240 & @xmath313 & @xmath242 & @xmath1270 & @xmath1304 & @xmath234 & @xmath1305 & @xmath1306 + ps111wj & @xmath453 & @xmath226 & @xmath1307 & @xmath1308 & @xmath240 & @xmath386 & @xmath242 & @xmath1163 & @xmath1309 & @xmath245 & @xmath1310 & @xmath362 + ps111yh & @xmath459 & @xmath226 & @xmath1311 & @xmath770 & @xmath240 & @xmath329 & @xmath231 & @xmath1312 & @xmath1313 & @xmath245 & @xmath1314 & @xmath1315 + ps111amh & @xmath463 & @xmath226 & @xmath1316 & @xmath1317 & @xmath240 & @xmath890 & @xmath242 & @xmath1149 & @xmath1318 & @xmath245 & @xmath1319 & @xmath1320 + ps111amv & @xmath471 & @xmath592 & @xmath1321 & @xmath1322 & @xmath240 & @xmath313 & @xmath262 & @xmath866 & @xmath1323 & @xmath234 & @xmath1324 & @xmath1325 + ps111anm & @xmath479 & @xmath226 & @xmath1326 & @xmath1327 & @xmath312 & @xmath1328 & @xmath262 & @xmath1133 & @xmath1329 & @xmath245 & @xmath1330 & @xmath1331 + ps111aof & @xmath485 & @xmath226 & @xmath1332 & @xmath552 & @xmath240 & @xmath1333 & @xmath231 & @xmath866 & @xmath1334 & @xmath245 & @xmath1335 & @xmath1336 + ps111apd & @xmath489 & @xmath226 & @xmath1337 & @xmath898 & @xmath1338 & @xmath1339 & @xmath428 & @xmath1133 & @xmath1340 & @xmath367 & @xmath1341 & @xmath1342 + ps111asz & @xmath497 & @xmath226 & @xmath1343 & @xmath1344 & @xmath648 & @xmath492 & @xmath231 & @xmath1345 & @xmath1346 & @xmath1347 & @xmath1348 & @xmath1349 + ps111awd & @xmath502 & @xmath226 & @xmath1350 & @xmath1351 & @xmath1352 & @xmath1353 & @xmath960 & @xmath1354 & @xmath1355 & @xmath367 & @xmath1356 & @xmath1357 + ps111azd & @xmath508 & @xmath226 & @xmath1358 & @xmath1359 & @xmath240 & @xmath890 & @xmath242 & @xmath1149 & @xmath1360 & @xmath245 & @xmath1361 & @xmath1362 + ps111azi & @xmath516 & @xmath226 & @xmath1363 & @xmath977 & @xmath953 & @xmath1364 & @xmath231 & @xmath1365 & @xmath1366 & @xmath623 & @xmath1367 & @xmath1368 + ps111azt & @xmath521 & @xmath226 & @xmath1369 & @xmath1370 & @xmath475 & @xmath1371 & @xmath262 & @xmath1133 & @xmath1372 & @xmath245 & @xmath1373 & @xmath1374 + ps111azv & @xmath528 & @xmath226 & @xmath1375 & @xmath928 & @xmath475 & @xmath1376 & @xmath428 & @xmath1377 & @xmath1378 & @xmath234 & @xmath1379 & @xmath823 + ps111bbg & @xmath532 & @xmath226 & @xmath1380 & @xmath770 & @xmath229 & @xmath540 & @xmath1381 & @xmath1382 & @xmath1383 & @xmath1384 & @xmath1385 & @xmath1386 + ps112c & @xmath538 & @xmath226 & @xmath1358 & @xmath1138 & @xmath1387 & @xmath1388 & @xmath242 & @xmath1245 & @xmath1389 & @xmath1390 & @xmath1391 & @xmath1392 + ps112d & @xmath545 & @xmath226 & @xmath1393 & @xmath898 & @xmath1394 & @xmath320 & @xmath231 & @xmath1245 & @xmath1395 & @xmath949 & @xmath1396 & @xmath1397 + ps112y & @xmath551 & @xmath226 & @xmath1159 & @xmath1398 & @xmath1399 & @xmath1400 & @xmath428 & @xmath1238 & @xmath1401 & @xmath1402 & @xmath1403 & @xmath1404 + ps112bt & @xmath559 & @xmath226 & @xmath1358 & @xmath1148 & @xmath240 & @xmath928 & @xmath262 & @xmath1238 & @xmath1405 & @xmath245 & @xmath1406 & @xmath342 + ps112rk & @xmath565 & @xmath226 & @xmath1407 & @xmath1408 & @xmath229 & @xmath1409 & @xmath262 & @xmath1127 & @xmath1410 & @xmath513 & @xmath1411 & @xmath1412 + ps112wn & @xmath572 & @xmath226 & @xmath947 & @xmath1413 & @xmath240 & @xmath321 & @xmath231 & @xmath1133 & @xmath1414 & @xmath234 & @xmath1415 & @xmath1416 + ps112zd & @xmath577 & @xmath226 & @xmath808 & @xmath336 & @xmath250 & @xmath1417 & @xmath242 & @xmath1418 & @xmath1419 & @xmath234 & @xmath1420 & @xmath1421 + ps112zp & @xmath582 & @xmath1422 & @xmath1423 & @xmath290 & @xmath229 & @xmath1174 & @xmath1424 & @xmath1425 & @xmath1426 & @xmath513 & @xmath1427 & @xmath798 + ps112auw & @xmath586 & @xmath226 & @xmath1428 & @xmath1429 & @xmath312 & @xmath952 & @xmath428 & @xmath1430 & @xmath1431 & @xmath669 & @xmath1432 & @xmath1433 + ps112bku & @xmath591 & @xmath226 & @xmath1434 & @xmath970 & @xmath1435 & @xmath783 & @xmath262 & @xmath1055 & @xmath1436 & @xmath245 & @xmath1437 & @xmath1438 + ps112bmp & @xmath600 & @xmath226 & @xmath1439 & @xmath1224 & @xmath298 & @xmath1440 & @xmath231 & @xmath1441 & @xmath1442 & @xmath1443 & @xmath1444 & @xmath1445 + ps112bml & @xmath607 & @xmath226 & @xmath1446 & @xmath498 & @xmath240 & @xmath1000 & @xmath322 & @xmath773 & @xmath1447 & @xmath245 & @xmath1448 & @xmath611 + ps112blz & @xmath612 & @xmath226 & @xmath1449 & @xmath594 & @xmath1450 & @xmath770 & @xmath322 & @xmath243 & @xmath1451 & @xmath1452 & @xmath1453 & @xmath1454 + ps112boi & @xmath619 & @xmath226 & @xmath1455 & @xmath290 & @xmath229 & @xmath1456 & @xmath262 & @xmath1312 & @xmath1457 & @xmath367 & @xmath1458 & @xmath1459 + ps112bza & @xmath625 & @xmath226 & @xmath1460 & @xmath970 & @xmath298 & @xmath561 & @xmath322 & @xmath243 & @xmath1461 & @xmath1462 & @xmath1463 & @xmath1464 + ps112cey & @xmath631 & @xmath1465 & @xmath1466 & @xmath259 & @xmath229 & @xmath1467 & @xmath262 & @xmath874 & @xmath1468 & @xmath245 & @xmath1469 & @xmath1470 + ps113hp & @xmath635 & @xmath226 & @xmath1471 & @xmath783 & @xmath229 & @xmath1472 & @xmath231 & @xmath1273 & @xmath1473 & @xmath949 & @xmath1474 & @xmath1475 + ps113vs & @xmath641 & @xmath226 & @xmath1476 & @xmath1477 & @xmath229 & @xmath1478 & @xmath231 & @xmath1345 & @xmath1479 & @xmath513 & @xmath1480 & @xmath1481 + ps113wr & @xmath646 & @xmath226 & @xmath1393 & @xmath1482 & @xmath403 & @xmath1483 & @xmath262 & @xmath1484 & @xmath1485 & @xmath1486 & @xmath717 & @xmath1487 + ps113alz & @xmath654 & @xmath226 & @xmath1488 & @xmath779 & @xmath240 & @xmath1489 & @xmath242 & @xmath1490 & @xmath1491 & @xmath234 & @xmath1492 & @xmath1493 + ps113arp & @xmath661 & @xmath226 & @xmath1494 & @xmath1115 & @xmath240 & @xmath329 & @xmath242 & @xmath1495 & @xmath1496 & @xmath245 & @xmath1497 & @xmath1498 + ps113ave & @xmath666 & @xmath226 & @xmath1499 & @xmath601 & @xmath1494 & @xmath1500 & @xmath262 & @xmath1501 & @xmath1502 & @xmath1503 & @xmath1504 & @xmath1505 + ps113atm & @xmath672 & @xmath226 & @xmath1506 & @xmath1507 & @xmath523 & @xmath349 & @xmath242 & @xmath1508 & @xmath1509 & @xmath730 & @xmath1510 & @xmath1511 + ps113bvb & @xmath681 & @xmath472 & @xmath1512 & @xmath1513 & @xmath1514 & @xmath1515 & @xmath262 & @xmath1238 & @xmath1516 & @xmath730 & @xmath1517 & @xmath1518 + ps113cqb & @xmath686 & @xmath226 & @xmath899 & @xmath1115 & @xmath475 & @xmath682 & @xmath242 & @xmath1519 & @xmath323 & @xmath245 & @xmath1520 & @xmath1521 + ps113cua & @xmath690 & @xmath226 & @xmath1522 & @xmath1523 & @xmath240 & @xmath1524 & @xmath428 & @xmath1133 & @xmath1525 & @xmath1526 & @xmath1527 & @xmath1528 + ps113eck & @xmath694 & @xmath226 & @xmath1529 & @xmath552 & @xmath240 & @xmath1530 & @xmath262 & @xmath1255 & @xmath263 & @xmath234 & @xmath1531 & @xmath1532 + ps113esn & @xmath698 & @xmath226 & @xmath1533 & @xmath397 & @xmath240 & @xmath313 & @xmath262 & @xmath1255 & @xmath1534 & @xmath245 & @xmath1535 & @xmath1536 + + ps110a & @xmath225 & @xmath226 & @xmath533 & @xmath465 & @xmath240 & @xmath627 & @xmath262 & @xmath1537 & @xmath1538 & @xmath245 & @xmath1539 & @xmath362 + ps110b & @xmath237 & @xmath226 & @xmath1540 & @xmath1541 & @xmath240 & @xmath329 & @xmath242 & @xmath1542 & @xmath1543 & @xmath245 & @xmath1544 & @xmath1545 + ps110q & @xmath248 & @xmath226 & @xmath1546 & @xmath1547 & @xmath921 & @xmath1548 & @xmath242 & @xmath1549 & @xmath1550 & @xmath1551 & @xmath1552 & @xmath1553 + ps110 t & @xmath257 & @xmath226 & @xmath1554 & @xmath1555 & @xmath636 & @xmath1328 & @xmath262 & @xmath1556 & @xmath1557 & @xmath245 & @xmath1558 & @xmath1559 + ps110ae & @xmath266 & @xmath226 & @xmath1103 & @xmath1560 & @xmath1561 & @xmath1562 & @xmath262 & @xmath1563 & @xmath1564 & @xmath367 & @xmath1565 & @xmath1566 + ps110lv & @xmath276 & @xmath226 & @xmath1567 & @xmath1568 & @xmath1569 & @xmath1570 & @xmath242 & @xmath1571 & @xmath1572 & @xmath367 & @xmath1573 & @xmath1574 + ps110 mb & @xmath283 & @xmath226 & @xmath1575 & @xmath1576 & @xmath291 & @xmath1577 & @xmath262 & @xmath1578 & @xmath1579 & @xmath949 & @xmath1580 & @xmath1581 + ps110zu & @xmath289 & @xmath226 & @xmath1582 & @xmath1583 & @xmath1584 & @xmath1585 & @xmath242 & @xmath1586 & @xmath1587 & @xmath1588 & @xmath1589 & @xmath1590 + ps110abd & @xmath297 & @xmath226 & @xmath1591 & @xmath1592 & @xmath229 & @xmath1593 & @xmath242 & @xmath1594 & @xmath1595 & @xmath730 & @xmath1596 & @xmath1597 + ps110agh & @xmath302 & @xmath226 & @xmath1598 & @xmath1599 & @xmath757 & @xmath1600 & @xmath231 & @xmath1601 & @xmath1602 & @xmath1603 & @xmath1604 & @xmath1605 + ps110awc & @xmath309 & @xmath226 & @xmath348 & @xmath1606 & @xmath953 & @xmath1607 & @xmath960 & @xmath1608 & @xmath1609 & @xmath924 & @xmath1610 & @xmath1611 + ps110biu & @xmath318 & @xmath226 & @xmath1612 & @xmath1613 & @xmath1614 & @xmath1476 & @xmath231 & @xmath1615 & @xmath1616 & @xmath623 & @xmath1617 & @xmath1618 + ps110blk & @xmath326 & @xmath1422 & @xmath1619 & @xmath1547 & @xmath240 & @xmath386 & @xmath1620 & @xmath1068 & @xmath435 & @xmath245 & @xmath1621 & @xmath1622 + ps111u & @xmath334 & @xmath226 & @xmath1623 & @xmath1624 & @xmath1625 & @xmath1626 & @xmath262 & @xmath1627 & @xmath1628 & @xmath1629 & @xmath1630 & @xmath342 + ps111z & @xmath343 & @xmath226 & @xmath1631 & @xmath1632 & @xmath1633 & @xmath1634 & @xmath231 & @xmath1635 & @xmath1636 & @xmath956 & @xmath1637 & @xmath1638 + ps111ai & @xmath347 & @xmath226 & @xmath1639 & @xmath1640 & @xmath229 & @xmath1641 & @xmath231 & @xmath1642 & @xmath1643 & @xmath513 & @xmath1644 & @xmath1645 + ps111ak & @xmath352 & @xmath226 & @xmath1646 & @xmath1647 & @xmath240 & @xmath890 & @xmath262 & @xmath1648 & @xmath1262 & @xmath234 & @xmath1649 & @xmath1650 + ps111bo & @xmath358 & @xmath226 & @xmath687 & @xmath1651 & @xmath240 & @xmath354 & @xmath242 & @xmath1652 & @xmath1653 & @xmath245 & @xmath1654 & @xmath1655 + ps111fl & @xmath363 & @xmath226 & @xmath1656 & @xmath1657 & @xmath1658 & @xmath1659 & @xmath242 & @xmath1660 & @xmath1661 & @xmath1526 & @xmath1662 & @xmath1663 + ps111jg & @xmath370 & @xmath226 & @xmath1623 & @xmath1664 & @xmath636 & @xmath1665 & @xmath231 & @xmath1666 & @xmath1667 & @xmath245 & @xmath1668 & @xmath693 + ps111kf & @xmath376 & @xmath226 & @xmath1669 & @xmath1670 & @xmath523 & @xmath313 & @xmath242 & @xmath1666 & @xmath1671 & @xmath234 & @xmath1672 & @xmath1673 + ps111li & @xmath384 & @xmath226 & @xmath1674 & @xmath1675 & @xmath229 & @xmath251 & @xmath231 & @xmath1676 & @xmath1677 & @xmath730 & @xmath1678 & @xmath1679 + ps111lj & @xmath390 & @xmath226 & @xmath1680 & @xmath364 & @xmath475 & @xmath1681 & @xmath242 & @xmath1682 & @xmath1683 & @xmath254 & @xmath1684 & @xmath1685 + ps111lr & @xmath396 & @xmath226 & @xmath1477 & @xmath1686 & @xmath240 & @xmath386 & @xmath231 & @xmath1687 & @xmath1688 & @xmath234 & @xmath1689 & @xmath938 + ps111lq & @xmath401 & @xmath226 & @xmath1690 & @xmath1691 & @xmath523 & @xmath911 & @xmath242 & @xmath1692 & @xmath1693 & @xmath234 & @xmath1694 & @xmath1695 + ps111md & @xmath408 & @xmath226 & @xmath1696 & @xmath1697 & @xmath240 & @xmath354 & @xmath231 & @xmath1698 & @xmath1699 & @xmath245 & @xmath1700 & @xmath282 + ps111pf & @xmath412 & @xmath226 & @xmath1701 & @xmath397 & @xmath475 & @xmath386 & @xmath242 & @xmath1702 & @xmath1703 & @xmath234 & @xmath1704 & @xmath1618 + ps111qe & @xmath417 & @xmath226 & @xmath1705 & @xmath259 & @xmath899 & @xmath890 & @xmath262 & @xmath1706 & @xmath476 & @xmath245 & @xmath1707 & @xmath1708 + ps111si & @xmath426 & @xmath226 & @xmath498 & @xmath1613 & @xmath240 & @xmath880 & @xmath262 & @xmath1709 & @xmath1710 & @xmath234 & @xmath1711 & @xmath1712 + ps111sm & @xmath432 & @xmath226 & @xmath517 & @xmath1713 & @xmath1561 & @xmath386 & @xmath242 & @xmath1714 & @xmath1715 & @xmath234 & @xmath1716 & @xmath1717 + ps111ua & @xmath438 & @xmath226 & @xmath1718 & @xmath348 & @xmath229 & @xmath386 & @xmath262 & @xmath1719 & @xmath1720 & @xmath234 & @xmath1721 & @xmath1722 + ps111uv & @xmath443 & @xmath226 & @xmath1723 & @xmath1724 & @xmath240 & @xmath329 & @xmath242 & @xmath1725 & @xmath633 & @xmath245 & @xmath1726 & @xmath693 + ps111vn & @xmath448 & @xmath592 & @xmath1727 & @xmath1728 & @xmath240 & @xmath261 & @xmath960 & @xmath1627 & @xmath1729 & @xmath245 & @xmath1730 & @xmath1731 + ps111wj & @xmath453 & @xmath226 & @xmath1732 & @xmath1733 & @xmath240 & @xmath386 & @xmath262 & @xmath1542 & @xmath1734 & @xmath234 & @xmath1735 & @xmath849 + ps111yh & @xmath459 & @xmath226 & @xmath1736 & @xmath1737 & @xmath475 & @xmath890 & @xmath262 & @xmath1738 & @xmath1739 & @xmath245 & @xmath1740 & @xmath1741 + ps111amh & @xmath463 & @xmath226 & @xmath1742 & @xmath1743 & @xmath240 & @xmath305 & @xmath242 & @xmath1542 & @xmath1744 & @xmath245 & @xmath1745 & @xmath1746 + ps111amv & @xmath471 & @xmath592 & @xmath1747 & @xmath1748 & @xmath240 & @xmath354 & @xmath242 & @xmath794 & @xmath663 & @xmath234 & @xmath1749 & @xmath846 + ps111anm & @xmath479 & @xmath226 & @xmath1750 & @xmath1751 & @xmath240 & @xmath285 & @xmath242 & @xmath1752 & @xmath1753 & @xmath245 & @xmath1754 & @xmath1755 + ps111aof & @xmath485 & @xmath226 & @xmath1756 & @xmath1757 & @xmath291 & @xmath1758 & @xmath242 & @xmath1759 & @xmath1760 & @xmath367 & @xmath1761 & @xmath1762 + ps111apd & @xmath489 & @xmath226 & @xmath1763 & @xmath1764 & @xmath934 & @xmath1765 & @xmath231 & @xmath1163 & @xmath1766 & @xmath1767 & @xmath1768 & @xmath1769 + ps111asz & @xmath497 & @xmath226 & @xmath1770 & @xmath1771 & @xmath420 & @xmath1772 & @xmath1424 & @xmath1047 & @xmath1773 & @xmath1774 & @xmath1775 & @xmath1776 + ps111awd & @xmath502 & @xmath226 & @xmath984 & @xmath1777 & @xmath757 & @xmath1778 & @xmath242 & @xmath1779 & @xmath1780 & @xmath245 & @xmath1781 & @xmath1782 + ps111azd & @xmath508 & @xmath1422 & @xmath1783 & @xmath1784 & @xmath636 & @xmath1697 & @xmath960 & @xmath1714 & @xmath1785 & @xmath234 & @xmath1786 & @xmath1787 + ps111azi & @xmath516 & @xmath226 & @xmath378 & @xmath1301 & @xmath1387 & @xmath1788 & @xmath262 & @xmath1280 & @xmath1789 & @xmath1790 & @xmath1791 & @xmath1673 + ps111azt & @xmath521 & @xmath226 & @xmath1792 & @xmath1793 & @xmath291 & @xmath1794 & @xmath1795 & @xmath1796 & @xmath1797 & @xmath494 & @xmath1798 & @xmath1799 + ps111azv & @xmath528 & @xmath226 & @xmath1800 & @xmath1801 & @xmath1802 & @xmath1803 & @xmath262 & @xmath1804 & @xmath1805 & @xmath1484 & @xmath1806 & @xmath1807 + ps111bbg & @xmath532 & @xmath226 & @xmath259 & @xmath977 & @xmath327 & @xmath1808 & @xmath242 & @xmath1809 & @xmath1810 & @xmath1811 & @xmath1812 & @xmath1813 + ps112c & @xmath538 & @xmath226 & @xmath1814 & @xmath1115 & @xmath450 & @xmath1815 & @xmath262 & @xmath1816 & @xmath1817 & @xmath1818 & @xmath1819 & @xmath1820 + ps112d & @xmath545 & @xmath226 & @xmath1821 & @xmath1822 & @xmath1823 & @xmath1824 & @xmath242 & @xmath1825 & @xmath1826 & @xmath1827 & @xmath1828 & @xmath1829 + ps112y & @xmath551 & @xmath226 & @xmath1830 & @xmath1831 & @xmath1832 & @xmath1833 & @xmath242 & @xmath1834 & @xmath1835 & @xmath1062 & @xmath1836 & @xmath645 + ps112bt & @xmath559 & @xmath226 & @xmath1837 & @xmath1838 & @xmath312 & @xmath1839 & @xmath428 & @xmath1666 & @xmath1840 & @xmath678 & @xmath1841 & @xmath909 + ps112rk & @xmath565 & @xmath226 & @xmath647 & @xmath1842 & @xmath229 & @xmath1843 & @xmath322 & @xmath1844 & @xmath1845 & @xmath367 & @xmath1846 & @xmath1847 + ps112wn & @xmath572 & @xmath226 & @xmath1499 & @xmath1848 & @xmath229 & @xmath1849 & @xmath242 & @xmath1676 & @xmath1850 & @xmath1384 & @xmath1851 & @xmath1852 + ps112zd & @xmath577 & @xmath592 & @xmath1853 & @xmath1792 & @xmath229 & @xmath547 & @xmath1424 & @xmath1854 & @xmath1329 & @xmath234 & @xmath1855 & @xmath1856 + ps112zp & @xmath582 & @xmath226 & @xmath1857 & @xmath1858 & @xmath250 & @xmath728 & @xmath242 & @xmath1859 & @xmath1860 & @xmath234 & @xmath1861 & @xmath798 + ps112auw & @xmath586 & @xmath226 & @xmath1664 & @xmath1862 & @xmath240 & @xmath1619 & @xmath231 & @xmath1863 & @xmath1864 & @xmath949 & @xmath1865 & @xmath1866 + ps112bku & @xmath591 & @xmath226 & @xmath419 & @xmath1867 & @xmath391 & @xmath1868 & @xmath428 & @xmath1869 & @xmath1870 & @xmath967 & @xmath1871 & @xmath798 + ps112bmp & @xmath600 & @xmath226 & @xmath984 & @xmath1872 & @xmath1873 & @xmath1874 & @xmath262 & @xmath1875 & @xmath1876 & @xmath1877 & @xmath1878 & @xmath1879 + ps112bml & @xmath607 & @xmath226 & @xmath1593 & @xmath699 & @xmath475 & @xmath321 & @xmath322 & @xmath866 & @xmath1880 & @xmath234 & @xmath1448 & @xmath1881 + ps112blz & @xmath612 & @xmath226 & @xmath378 & @xmath984 & @xmath1882 & @xmath1169 & @xmath322 & @xmath941 & @xmath1883 & @xmath1884 & @xmath1885 & @xmath1886 + ps112boi & @xmath619 & @xmath226 & @xmath1887 & @xmath1888 & @xmath240 & @xmath434 & @xmath262 & @xmath1181 & @xmath1889 & @xmath245 & @xmath1890 & @xmath1891 + ps112bza & @xmath625 & @xmath226 & @xmath1892 & @xmath1893 & @xmath298 & @xmath1344 & @xmath242 & @xmath1894 & @xmath1895 & @xmath1896 & @xmath1897 & @xmath1898 + ps112cey & @xmath631 & @xmath592 & @xmath1899 & @xmath647 & @xmath1243 & @xmath1900 & @xmath231 & @xmath1273 & @xmath1901 & @xmath1902 & @xmath1903 & @xmath798 + ps113hp & @xmath635 & @xmath226 & @xmath1115 & @xmath779 & @xmath899 & @xmath1904 & @xmath242 & @xmath1905 & @xmath1906 & @xmath331 & @xmath1907 & @xmath1908 + ps113vs & @xmath641 & @xmath226 & @xmath1909 & @xmath1910 & @xmath229 & @xmath1690 & @xmath242 & @xmath1911 & @xmath1912 & @xmath949 & @xmath1913 & @xmath1914 + ps113wr & @xmath646 & @xmath226 & @xmath1915 & @xmath1916 & @xmath439 & @xmath1917 & @xmath231 & @xmath1648 & @xmath1918 & @xmath1919 & @xmath1920 & @xmath296 + ps113alz & @xmath654 & @xmath1921 & @xmath1922 & @xmath534 & @xmath229 & @xmath1923 & @xmath1424 & @xmath1924 & @xmath1925 & @xmath1384 & @xmath1926 & @xmath1927 + ps113arp & @xmath661 & @xmath226 & @xmath1928 & @xmath1842 & @xmath229 & @xmath1929 & @xmath242 & @xmath1930 & @xmath1931 & @xmath623 & @xmath1932 & @xmath1933 + ps113ave & @xmath666 & @xmath592 & @xmath348 & @xmath1934 & @xmath1935 & @xmath349 & @xmath960 & @xmath1936 & @xmath1937 & @xmath1938 & @xmath1939 & @xmath1940 + ps113atm & @xmath672 & @xmath226 & @xmath1941 & @xmath1942 & @xmath291 & @xmath1943 & @xmath242 & @xmath1944 & @xmath1945 & @xmath1062 & @xmath1946 & @xmath1362 + ps113bvb & @xmath681 & @xmath226 & @xmath1947 & @xmath1948 & @xmath1873 & @xmath1949 & @xmath231 & @xmath1950 & @xmath1951 & @xmath1952 & @xmath1953 & @xmath1954 + ps113cqb & @xmath686 & @xmath226 & @xmath1955 & @xmath1778 & @xmath240 & @xmath890 & @xmath262 & @xmath1068 & @xmath1956 & @xmath245 & @xmath1957 & @xmath1958 + ps113cua & @xmath690 & @xmath226 & @xmath397 & @xmath977 & @xmath240 & @xmath1097 & @xmath242 & @xmath1959 & @xmath1960 & @xmath254 & @xmath1961 & @xmath1962 + ps113eck & @xmath694 & @xmath226 & @xmath1963 & @xmath239 & @xmath523 & @xmath386 & @xmath242 & @xmath1964 & @xmath512 & @xmath245 & @xmath1965 & @xmath1966 + ps113esn & @xmath698 & @xmath226 & @xmath1967 & @xmath1968 & @xmath229 & @xmath1530 & @xmath242 & @xmath1698 & @xmath1969 & @xmath245 & @xmath1970 & @xmath645 + + ps110a & @xmath225 & @xmath226 & @xmath1971 & @xmath1972 & @xmath250 & @xmath329 & @xmath262 & @xmath1973 & @xmath1974 & @xmath234 & @xmath1975 & @xmath1545 + ps110b & @xmath237 & @xmath226 & @xmath1976 & @xmath1977 & @xmath240 & @xmath329 & @xmath242 & @xmath1978 & @xmath1979 & @xmath245 & @xmath1980 & @xmath1981 + ps110q & @xmath248 & @xmath226 & @xmath1982 & @xmath1983 & @xmath240 & @xmath329 & @xmath262 & @xmath1984 & @xmath881 & @xmath234 & @xmath1985 & @xmath1986 + ps110 t & @xmath257 & @xmath226 & @xmath1987 & @xmath1972 & @xmath240 & @xmath313 & @xmath262 & @xmath1988 & @xmath1979 & @xmath245 & @xmath1989 & @xmath1990 + ps110ae & @xmath266 & @xmath226 & @xmath1991 & @xmath1992 & @xmath240 & @xmath386 & @xmath262 & @xmath1988 & @xmath1993 & @xmath245 & @xmath1994 & @xmath1995 + ps110lv & @xmath276 & @xmath226 & @xmath1996 & @xmath1997 & @xmath240 & @xmath329 & @xmath231 & @xmath1988 & @xmath1688 & @xmath234 & @xmath1998 & @xmath1999 + ps110 mb & @xmath283 & @xmath226 & @xmath2000 & @xmath1977 & @xmath240 & @xmath329 & @xmath262 & @xmath1978 & @xmath2001 & @xmath234 & @xmath2002 & @xmath1999 + ps110zu & @xmath289 & @xmath226 & @xmath2003 & @xmath2004 & @xmath475 & @xmath890 & @xmath262 & @xmath1984 & @xmath1912 & @xmath234 & @xmath2005 & @xmath1722 + ps110abd & @xmath297 & @xmath226 & @xmath2006 & @xmath2007 & @xmath240 & @xmath329 & @xmath428 & @xmath1984 & @xmath1974 & @xmath234 & @xmath2008 & @xmath1999 + ps110agh & @xmath302 & @xmath226 & @xmath2009 & @xmath2010 & @xmath240 & @xmath1530 & @xmath262 & @xmath1988 & @xmath2011 & @xmath245 & @xmath2012 & @xmath2013 + ps110awc & @xmath309 & @xmath226 & @xmath2014 & @xmath2007 & @xmath240 & @xmath386 & @xmath242 & @xmath1973 & @xmath1703 & @xmath234 & @xmath2015 & @xmath375 + ps110biu & @xmath318 & @xmath226 & @xmath2016 & @xmath2017 & @xmath240 & @xmath305 & @xmath262 & @xmath1984 & @xmath2018 & @xmath245 & @xmath2019 & @xmath2013 + ps110blk & @xmath326 & @xmath226 & @xmath2020 & @xmath2021 & @xmath240 & @xmath404 & @xmath262 & @xmath1988 & @xmath1496 & @xmath245 & @xmath2022 & @xmath1995 + ps111u & @xmath334 & @xmath226 & @xmath2023 & @xmath1983 & @xmath240 & @xmath313 & @xmath262 & @xmath2024 & @xmath2025 & @xmath234 & @xmath2026 & @xmath2027 + ps111z & @xmath343 & @xmath226 & @xmath2028 & @xmath2029 & @xmath240 & @xmath1097 & @xmath231 & @xmath2030 & @xmath2031 & @xmath245 & @xmath2032 & @xmath1990 + ps111ai & @xmath347 & @xmath226 & @xmath2028 & @xmath2033 & @xmath240 & @xmath329 & @xmath231 & @xmath1984 & @xmath2034 & @xmath245 & @xmath2035 & @xmath375 + ps111ak & @xmath352 & @xmath226 & @xmath2036 & @xmath1983 & @xmath240 & @xmath251 & @xmath231 & @xmath1984 & @xmath695 & @xmath245 & @xmath2037 & @xmath2038 + ps111bo & @xmath358 & @xmath226 & @xmath2039 & @xmath2040 & @xmath240 & @xmath404 & @xmath262 & @xmath1988 & @xmath2041 & @xmath234 & @xmath2042 & @xmath2043 + ps111fl & @xmath363 & @xmath226 & @xmath2028 & @xmath2044 & @xmath523 & @xmath329 & @xmath231 & @xmath1936 & @xmath2045 & @xmath254 & @xmath2046 & @xmath2047 + ps111jg & @xmath370 & @xmath226 & @xmath2048 & @xmath2049 & @xmath240 & @xmath329 & @xmath242 & @xmath1984 & @xmath844 & @xmath245 & @xmath2050 & @xmath2051 + ps111kf & @xmath376 & @xmath226 & @xmath2052 & @xmath1983 & @xmath240 & @xmath305 & @xmath231 & @xmath2053 & @xmath2054 & @xmath245 & @xmath2055 & @xmath2056 + ps111li & @xmath384 & @xmath226 & @xmath2057 & @xmath1997 & @xmath240 & @xmath329 & @xmath262 & @xmath2058 & @xmath476 & @xmath245 & @xmath2059 & @xmath2060 + ps111lj & @xmath390 & @xmath226 & @xmath2006 & @xmath2061 & @xmath915 & @xmath241 & @xmath242 & @xmath1973 & @xmath1974 & @xmath245 & @xmath2062 & @xmath2063 + ps111lr & @xmath396 & @xmath226 & @xmath2064 & @xmath2065 & @xmath240 & @xmath404 & @xmath242 & @xmath2066 & @xmath2067 & @xmath245 & @xmath2068 & @xmath2069 + ps111lq & @xmath401 & @xmath226 & @xmath2070 & @xmath2071 & @xmath636 & @xmath354 & @xmath242 & @xmath1988 & @xmath2072 & @xmath245 & @xmath2073 & @xmath2074 + ps111md & @xmath408 & @xmath226 & @xmath2028 & @xmath1972 & @xmath240 & @xmath329 & @xmath242 & @xmath1973 & @xmath2075 & @xmath245 & @xmath2076 & @xmath1545 + ps111pf & @xmath412 & @xmath226 & @xmath2077 & @xmath2078 & @xmath240 & @xmath404 & @xmath428 & @xmath1973 & @xmath1309 & @xmath245 & @xmath2079 & @xmath2080 + ps111qe & @xmath417 & @xmath226 & @xmath2081 & @xmath1972 & @xmath240 & @xmath2082 & @xmath242 & @xmath1988 & @xmath2083 & @xmath245 & @xmath2084 & @xmath2085 + ps111si & @xmath426 & @xmath226 & @xmath2086 & @xmath1977 & @xmath899 & @xmath2087 & @xmath242 & @xmath1978 & @xmath709 & @xmath234 & @xmath2088 & @xmath2085 + ps111sm & @xmath432 & @xmath226 & @xmath2020 & @xmath1972 & @xmath240 & @xmath305 & @xmath231 & @xmath1984 & @xmath1739 & @xmath234 & @xmath2089 & @xmath375 + ps111ua & @xmath438 & @xmath226 & @xmath2077 & @xmath2044 & @xmath240 & @xmath241 & @xmath262 & @xmath2090 & @xmath2091 & @xmath245 & @xmath2092 & @xmath2093 + ps111uv & @xmath443 & @xmath226 & @xmath2094 & @xmath2095 & @xmath240 & @xmath354 & @xmath242 & @xmath1984 & @xmath2096 & @xmath245 & @xmath2097 & @xmath1362 + ps111vn & @xmath448 & @xmath226 & @xmath2098 & @xmath2099 & @xmath240 & @xmath890 & @xmath242 & @xmath1978 & @xmath2100 & @xmath234 & @xmath2101 & @xmath425 + ps111wj & @xmath453 & @xmath226 & @xmath2077 & @xmath2029 & @xmath240 & @xmath329 & @xmath242 & @xmath1978 & @xmath2102 & @xmath245 & @xmath2103 & @xmath1981 + ps111yh & @xmath459 & @xmath226 & @xmath2064 & @xmath2007 & @xmath240 & @xmath329 & @xmath231 & @xmath2030 & @xmath2104 & @xmath245 & @xmath2105 & @xmath375 + ps111amh & @xmath463 & @xmath226 & @xmath2106 & @xmath2107 & @xmath240 & @xmath662 & @xmath231 & @xmath2108 & @xmath2034 & @xmath245 & @xmath2109 & @xmath2110 + ps111amv & @xmath471 & @xmath226 & @xmath2111 & @xmath2112 & @xmath240 & @xmath313 & @xmath262 & @xmath2053 & @xmath2113 & @xmath245 & @xmath2114 & @xmath693 + ps111anm & @xmath479 & @xmath226 & @xmath2115 & @xmath421 & @xmath240 & @xmath285 & @xmath262 & @xmath1973 & @xmath1262 & @xmath245 & @xmath2116 & @xmath2117 + ps111aof & @xmath485 & @xmath226 & @xmath2118 & @xmath994 & @xmath899 & @xmath682 & @xmath242 & @xmath2119 & @xmath2120 & @xmath234 & @xmath2121 & @xmath2122 + ps111apd & @xmath489 & @xmath226 & @xmath2123 & @xmath2124 & @xmath2125 & @xmath889 & @xmath231 & @xmath2126 & @xmath2127 & @xmath725 & @xmath2128 & @xmath2129 + ps111asz & @xmath497 & @xmath226 & @xmath2130 & @xmath2131 & @xmath240 & @xmath1833 & @xmath242 & @xmath2132 & @xmath2075 & @xmath245 & @xmath2133 & @xmath2134 + ps111awd & @xmath502 & @xmath226 & @xmath2135 & @xmath2136 & @xmath757 & @xmath2137 & @xmath231 & @xmath2138 & @xmath2139 & @xmath245 & @xmath2140 & @xmath2141 + ps111azd & @xmath508 & @xmath226 & @xmath2142 & @xmath2143 & @xmath312 & @xmath2144 & @xmath242 & @xmath2145 & @xmath2146 & @xmath245 & @xmath2147 & @xmath2148 + ps111azi & @xmath516 & @xmath226 & @xmath1997 & @xmath1450 & @xmath229 & @xmath1417 & @xmath242 & @xmath2149 & @xmath2150 & @xmath245 & @xmath2151 & @xmath1362 + ps111azt & @xmath521 & @xmath226 & @xmath2152 & @xmath2153 & @xmath240 & @xmath386 & @xmath262 & @xmath2145 & @xmath2154 & @xmath245 & @xmath2155 & @xmath2156 + ps111azv & @xmath528 & @xmath226 & @xmath2157 & @xmath2158 & @xmath240 & @xmath278 & @xmath231 & @xmath2159 & @xmath2160 & @xmath245 & @xmath2161 & @xmath325 + ps111bbg & @xmath532 & @xmath226 & @xmath2162 & @xmath2163 & @xmath229 & @xmath2164 & @xmath242 & @xmath2132 & @xmath2165 & @xmath949 & @xmath2166 & @xmath247 + ps112c & @xmath538 & @xmath226 & @xmath2167 & @xmath2168 & @xmath240 & @xmath313 & @xmath262 & @xmath2058 & @xmath2169 & @xmath234 & @xmath2170 & @xmath2171 + ps112d & @xmath545 & @xmath226 & @xmath2172 & @xmath2173 & @xmath721 & @xmath2174 & @xmath231 & @xmath2175 & @xmath2176 & @xmath725 & @xmath2177 & @xmath1299 + ps112y & @xmath551 & @xmath226 & @xmath2178 & @xmath2179 & @xmath229 & @xmath1332 & @xmath231 & @xmath2180 & @xmath663 & @xmath730 & @xmath2181 & @xmath2182 + ps112bt & @xmath559 & @xmath226 & @xmath2183 & @xmath2184 & @xmath636 & @xmath2185 & @xmath231 & @xmath2186 & @xmath2187 & @xmath669 & @xmath2188 & @xmath1866 + ps112rk & @xmath565 & @xmath226 & @xmath2099 & @xmath2189 & @xmath1561 & @xmath2190 & @xmath262 & @xmath2191 & @xmath286 & @xmath1384 & @xmath2192 & @xmath2193 + ps112wn & @xmath572 & @xmath226 & @xmath2194 & @xmath649 & @xmath240 & @xmath2195 & @xmath242 & @xmath2196 & @xmath1729 & @xmath730 & @xmath2197 & @xmath2198 + ps112zd & @xmath577 & @xmath226 & @xmath2199 & @xmath2200 & @xmath229 & @xmath534 & @xmath242 & @xmath1988 & @xmath1693 & @xmath513 & @xmath2201 & @xmath2202 + ps112zp & @xmath582 & @xmath226 & @xmath2203 & @xmath2204 & @xmath240 & @xmath1326 & @xmath242 & @xmath2205 & @xmath2100 & @xmath367 & @xmath2206 & @xmath247 + ps112auw & @xmath586 & @xmath226 & @xmath2207 & @xmath2208 & @xmath229 & @xmath2209 & @xmath723 & @xmath2053 & @xmath1960 & @xmath367 & @xmath2210 & @xmath2211 + ps112bku & @xmath591 & @xmath1422 & @xmath2212 & @xmath1000 & @xmath953 & @xmath2213 & @xmath1620 & @xmath2214 & @xmath2215 & @xmath274 & @xmath2216 & @xmath325 + ps112bmp & @xmath600 & @xmath226 & @xmath2217 & @xmath2218 & @xmath953 & @xmath2219 & @xmath231 & @xmath2066 & @xmath2220 & @xmath1297 & @xmath2221 & @xmath2222 + ps112bml & @xmath607 & @xmath226 & @xmath2223 & @xmath2224 & @xmath240 & @xmath313 & @xmath322 & @xmath2225 & @xmath1974 & @xmath245 & @xmath2226 & @xmath2227 + ps112blz & @xmath612 & @xmath226 & @xmath2228 & @xmath1169 & @xmath240 & @xmath2229 & @xmath262 & @xmath2230 & @xmath2231 & @xmath367 & @xmath2232 & @xmath2233 + ps112boi & @xmath619 & @xmath226 & @xmath1992 & @xmath2234 & @xmath240 & @xmath1187 & @xmath262 & @xmath2235 & @xmath2236 & @xmath949 & @xmath2237 & @xmath2238 + ps112bza & @xmath625 & @xmath226 & @xmath2239 & @xmath2240 & @xmath614 & @xmath2241 & @xmath262 & @xmath2242 & @xmath2243 & @xmath245 & @xmath2244 & @xmath2245 + ps112cey & @xmath631 & @xmath226 & @xmath2246 & @xmath1337 & @xmath260 & @xmath2247 & @xmath231 & @xmath2248 & @xmath2249 & @xmath815 & @xmath2250 & @xmath2251 + ps113hp & @xmath635 & @xmath226 & @xmath2252 & @xmath2253 & @xmath229 & @xmath2254 & @xmath428 & @xmath2196 & @xmath2255 & @xmath1384 & @xmath2256 & @xmath1981 + ps113vs & @xmath641 & @xmath226 & @xmath2257 & @xmath2258 & @xmath250 & @xmath1849 & @xmath231 & @xmath2259 & @xmath2260 & @xmath367 & @xmath2261 & @xmath2262 + ps113wr & @xmath646 & @xmath226 & @xmath2263 & @xmath1115 & @xmath260 & @xmath313 & @xmath262 & @xmath2264 & @xmath2265 & @xmath234 & @xmath2266 & @xmath2267 + ps113alz & @xmath654 & @xmath226 & @xmath2268 & @xmath728 & @xmath229 & @xmath241 & @xmath242 & @xmath2145 & @xmath2269 & @xmath1384 & @xmath2270 & @xmath2271 + ps113arp & @xmath661 & @xmath226 & @xmath2272 & @xmath2273 & @xmath240 & @xmath285 & @xmath262 & @xmath1988 & @xmath1534 & @xmath245 & @xmath2274 & @xmath2275 + ps113ave & @xmath666 & @xmath226 & @xmath1147 & @xmath2276 & @xmath475 & @xmath404 & @xmath242 & @xmath2277 & @xmath633 & @xmath234 & @xmath2278 & @xmath2279 + ps113atm & @xmath672 & @xmath226 & @xmath2280 & @xmath2281 & @xmath475 & @xmath2282 & @xmath428 & @xmath2283 & @xmath1785 & @xmath563 & @xmath2284 & @xmath2285 + ps113bvb & @xmath681 & @xmath226 & @xmath2286 & @xmath2287 & @xmath229 & @xmath854 & @xmath262 & @xmath2288 & @xmath2289 & @xmath513 & @xmath2290 & @xmath2291 + ps113cqb & @xmath686 & @xmath226 & @xmath2292 & @xmath2293 & @xmath240 & @xmath890 & @xmath231 & @xmath2294 & @xmath2295 & @xmath912 & @xmath2296 & @xmath2297 + ps113cua & @xmath690 & @xmath226 & @xmath2298 & @xmath2299 & @xmath268 & @xmath2300 & @xmath262 & @xmath2030 & @xmath2301 & @xmath730 & @xmath2302 & @xmath317 + ps113eck & @xmath694 & @xmath267 & @xmath2303 & @xmath2304 & @xmath240 & @xmath2305 & @xmath901 & @xmath2306 & @xmath2307 & @xmath730 & @xmath2308 & @xmath2309 + ps113esn & @xmath698 & @xmath226 & @xmath2310 & @xmath2065 & @xmath240 & @xmath386 & @xmath231 & @xmath1978 & @xmath1797 & @xmath245 & @xmath2311 & @xmath2047 lrrrrr -15 & @xmath2312 & @xmath2312 & @xmath2312 & @xmath2312 & @xmath2312 + -12 & @xmath2312 & @xmath2312 & @xmath2312 & @xmath2313 & @xmath2312 + -9 & @xmath2314 & @xmath2315 & @xmath2316 & @xmath2317 & @xmath2318 + -6 & @xmath2319 & @xmath2320 & @xmath2321 & @xmath2322 & @xmath2323 + -3 & @xmath2324 & @xmath2325 & @xmath2326 & @xmath2327 & @xmath2328 + 0 & @xmath2329 & @xmath2330 & @xmath2331 & @xmath2332 & @xmath2333 + 3 & @xmath2334 & @xmath2335 & @xmath2336 & @xmath2337 & @xmath2338 + 6 & @xmath2339 & @xmath2340 & @xmath2341 & @xmath2342 & @xmath2343 + 9 & @xmath2344 & @xmath2345 & @xmath2346 & @xmath2347 & @xmath2348 + 12 & @xmath2349 & @xmath2350 & @xmath2351 & @xmath2352 & @xmath2353 + 15 & @xmath2354 & @xmath2355 & @xmath2356 & @xmath2357 & @xmath2358 + 18 & @xmath2359 & @xmath2360 & @xmath2361 & @xmath2362 & @xmath2363 + 21 & @xmath2364 & @xmath2365 & @xmath2366 & @xmath2367 & @xmath2368 + 24 & @xmath2369 & @xmath2370 & @xmath2371 & @xmath2372 & @xmath2373 + 27 & @xmath2374 & @xmath2375 & @xmath2376 & @xmath2377 & @xmath2378 + 30 & @xmath2379 & @xmath2380 & @xmath2381 & @xmath2382 & @xmath2383 + 33 & @xmath2384 & @xmath2385 & @xmath2386 & @xmath2387 & @xmath2388 + 36 & @xmath2389 & @xmath2390 & @xmath2391 & @xmath2392 & @xmath2393 + 39 & @xmath2394 & @xmath2395 & @xmath2396 & @xmath2397 & @xmath2398 + 42 & @xmath2399 & @xmath2400 & @xmath2401 & @xmath2402 & @xmath2403 + 45 & @xmath2404 & @xmath2405 & @xmath2406 & @xmath2407 & @xmath2408 + 48 & @xmath2409 & @xmath2410 & @xmath2411 & @xmath2412 & @xmath2413 + 51 & @xmath2414 & @xmath2415 & @xmath2416 & @xmath2417 & @xmath2418 + 54 & @xmath2419 & @xmath2420 & @xmath2421 & @xmath2422 & @xmath2423 + 57 & @xmath2424 & @xmath2425 & @xmath2426 & @xmath2427 & @xmath2428 + 60 & @xmath2429 & @xmath2430 & @xmath2431 & @xmath2432 & @xmath2433 + 63 & @xmath2434 & @xmath2435 & @xmath2436 & @xmath2437 & @xmath2438 + 66 & @xmath2439 & @xmath2440 & @xmath2441 & @xmath2442 & @xmath2443 + 69 & @xmath2444 & @xmath2445 & @xmath2446 & @xmath2447 & @xmath2448 + 72 & @xmath2449 & @xmath2450 & @xmath2451 & @xmath2452 & @xmath2453 + 75 & @xmath2454 & @xmath2455 & @xmath2456 & @xmath2457 & @xmath2458 + 78 & @xmath2459 & @xmath2460 & @xmath2461 & @xmath2462 & @xmath2463 + 81 & @xmath2464 & @xmath2465 & @xmath2466 & @xmath2467 & @xmath2468 + 84 & @xmath2469 & @xmath2470 & @xmath2471 & @xmath2472 & @xmath2473 + 87 & @xmath2474 & @xmath2475 & @xmath2476 & @xmath2477 & @xmath2478 + 90 & @xmath2479 & @xmath2480 & @xmath2481 & @xmath2482 & @xmath2483 + 93 & @xmath2484 & @xmath2485 & @xmath2486 & @xmath2487 & @xmath2488 + 96 & @xmath2489 & @xmath2490 & @xmath2491 & @xmath2492 & @xmath2493 + 99 & @xmath2494 & @xmath2495 & @xmath2496 & @xmath2497 & @xmath2498 + 102 & @xmath2499 & @xmath2500 & @xmath2501 & @xmath2502 & @xmath2503 + 105 & @xmath2312 & @xmath2504 & @xmath2505 & @xmath2506 & @xmath2507 + 108 & @xmath2312 & @xmath2508 & @xmath2509 & @xmath2510 & @xmath2511 + 111 & @xmath2312 & @xmath2512 & @xmath2513 & @xmath2514 & @xmath2515 + 114 & @xmath2312 & @xmath2312 & @xmath2516 & @xmath2517 & @xmath2312 lllllllll ps110a & @xmath2518 & @xmath2519 & @xmath2520 & @xmath2521 & @xmath2522 & & & + ps110q & @xmath2523 & @xmath2524 & @xmath2525 & @xmath2526 & @xmath2527 & & & + ps110ae & @xmath2528 & @xmath2529 & @xmath2530 & @xmath2531 & @xmath2532 & & & + ps110lv & @xmath2533 & @xmath2534 & @xmath2535 & @xmath2536 & @xmath2537 & & & + ps110 mb & @xmath2538 & @xmath2539 & @xmath2540 & @xmath2541 & @xmath2542 & & & + ps110zu & @xmath2543 & @xmath2544 & @xmath2545 & @xmath2546 & @xmath2547 & & & + ps110abd & @xmath2548 & @xmath2549 & @xmath2550 & @xmath2551 & @xmath2552 & & & + ps110agh & @xmath2553 & @xmath2554 & @xmath2555 & @xmath2556 & @xmath2557 & @xmath2558 & @xmath2559 & @xmath2560 + ps110awc & @xmath2561 & @xmath2562 & @xmath2563 & @xmath2564 & @xmath2565 & & & + ps110biu & @xmath2566 & @xmath2567 & @xmath2568 & @xmath2569 & @xmath2570 & & & + ps110blk & @xmath2571 & @xmath2572 & @xmath2573 & @xmath2574 & @xmath2575 & & & + ps111u & @xmath2576 & @xmath2577 & @xmath2578 & @xmath2579 & @xmath2580 & & & + ps111z & @xmath2581 & @xmath2582 & @xmath2583 & @xmath2584 & @xmath2585 & & & + ps111ai & @xmath2586 & @xmath2587 & @xmath2588 & @xmath2589 & @xmath2590 & & & + ps111bo & @xmath2591 & @xmath2592 & @xmath2593 & @xmath2594 & @xmath2595 & & & + ps111fl & @xmath2596 & @xmath2597 & @xmath2598 & @xmath2599 & @xmath2600 & & & + ps111kf & @xmath2601 & @xmath2602 & @xmath2603 & @xmath2604 & @xmath2605 & @xmath2606 & @xmath2607 & @xmath2608 + ps111li & @xmath2609 & @xmath2610 & @xmath2611 & @xmath2612 & @xmath2613 & @xmath2614 & @xmath2615 & @xmath2616 + ps111lj & @xmath2617 & @xmath2618 & @xmath2619 & @xmath2620 & @xmath2621 & & & + ps111lr & @xmath2622 & @xmath2623 & @xmath2624 & @xmath2625 & @xmath2626 & & & + ps111lq & @xmath2627 & @xmath2628 & @xmath2629 & @xmath2630 & @xmath2631 & & & + ps111md & @xmath2632 & @xmath2633 & @xmath2634 & @xmath2635 & @xmath2636 & & & + ps111pf & @xmath2637 & @xmath2638 & @xmath2639 & @xmath2640 & @xmath2641 & & & + ps111qe & @xmath2642 & @xmath2643 & @xmath2644 & @xmath2645 & @xmath2646 & & & + ps111sm & @xmath2647 & @xmath2648 & @xmath2649 & @xmath2650 & @xmath2651 & & & + ps111ua & @xmath2652 & @xmath2653 & @xmath2654 & @xmath2655 & @xmath2656 & & & + ps111uv & @xmath2657 & @xmath2658 & @xmath2659 & @xmath2660 & @xmath2661 & & & + ps111vn & @xmath2662 & @xmath2663 & @xmath2664 & @xmath2665 & @xmath2666 & & & + ps111wj & @xmath2667 & @xmath2668 & @xmath2669 & @xmath2670 & @xmath2671 & & & + ps111yh & @xmath2672 & @xmath2673 & @xmath2674 & @xmath2675 & @xmath2676 & & & + ps111amh & @xmath2677 & @xmath2678 & @xmath2679 & @xmath2680 & @xmath2681 & & & + ps111anm & @xmath2682 & @xmath2683 & @xmath2684 & @xmath2685 & @xmath2686 & & & + ps111aof & @xmath2687 & @xmath2688 & @xmath2689 & @xmath2690 & @xmath2691 & @xmath2692 & @xmath2693 & @xmath2694 + ps111apd & @xmath2695 & @xmath2696 & @xmath2697 & @xmath2698 & @xmath2699 & & & + ps111asz & @xmath2700 & @xmath2701 & @xmath2702 & @xmath2703 & @xmath2704 & & & + ps111awd & @xmath2705 & @xmath2706 & @xmath2707 & @xmath2708 & @xmath2709 & & & + ps111azd & @xmath2710 & @xmath2711 & @xmath2712 & @xmath2713 & @xmath2714 & @xmath2715 & @xmath2716 & @xmath2717 + ps111azi & @xmath2718 & @xmath2719 & @xmath2720 & @xmath2721 & @xmath2722 & & & + ps111azt & @xmath2723 & @xmath2724 & @xmath2725 & @xmath2726 & @xmath2727 & & & + ps111azv & @xmath2728 & @xmath2729 & @xmath2730 & @xmath2731 & @xmath2732 & & & + ps111bbg & @xmath2733 & @xmath2734 & @xmath2735 & @xmath2736 & @xmath2737 & @xmath2738 & @xmath2739 & @xmath2740 + ps112c & @xmath2741 & @xmath2742 & @xmath2743 & @xmath2744 & @xmath2745 & & & + ps112d & @xmath2746 & @xmath2747 & @xmath2748 & @xmath2749 & @xmath2750 & & & + ps112y & @xmath2751 & @xmath2752 & @xmath2753 & @xmath2754 & @xmath2755 & & & + ps112rk & @xmath2756 & @xmath2757 & @xmath2758 & @xmath2759 & @xmath2760 & & & + ps112zd & @xmath2761 & @xmath2762 & @xmath2763 & @xmath2764 & @xmath2765 & & & + ps112zp & @xmath2766 & @xmath381 & @xmath2767 & @xmath2768 & @xmath2769 & & & + ps112auw & @xmath2770 & @xmath2771 & @xmath2772 & @xmath2773 & @xmath2774 & & & + ps112bku & @xmath2775 & @xmath2776 & @xmath2777 & @xmath2778 & @xmath2779 & & & + ps112bmp & @xmath2780 & @xmath2781 & @xmath2782 & @xmath2783 & @xmath2784 & & & + ps112blz & @xmath2785 & @xmath2786 & @xmath2787 & @xmath2788 & @xmath2789 & & & + ps112boi & @xmath2790 & @xmath2791 & @xmath2792 & @xmath2793 & @xmath2794 & & & + ps112bza & @xmath2795 & @xmath2796 & @xmath2797 & @xmath2798 & @xmath2799 & & & + ps112cey & @xmath2800 & @xmath2801 & @xmath2802 & @xmath2803 & @xmath2804 & & & + ps113hp & @xmath2805 & @xmath2806 & @xmath2807 & @xmath2630 & @xmath2808 & @xmath2809 & @xmath2810 & @xmath2811 + ps113vs & @xmath2812 & @xmath2813 & @xmath2814 & @xmath2815 & @xmath2816 & & & + ps113wr & @xmath2817 & @xmath2818 & @xmath2819 & @xmath2546 & @xmath2820 & & & + ps113alz & @xmath2821 & @xmath2822 & @xmath2823 & @xmath2824 & @xmath2825 & @xmath2826 & @xmath2827 & @xmath2828 + ps113arp & @xmath2829 & @xmath2830 & @xmath2831 & @xmath2832 & @xmath2833 & & & + ps113ave & @xmath2834 & @xmath2835 & @xmath2836 & @xmath2837 & @xmath2838 & & & + ps113atm & @xmath2839 & @xmath2840 & @xmath2841 & @xmath2842 & @xmath2843 & & & + ps113bvb & @xmath2844 & @xmath2845 & @xmath2846 & @xmath2847 & @xmath2848 & & & + ps113cqb & @xmath2849 & @xmath2850 & @xmath2851 & @xmath2852 & @xmath2853 & & & + ps113cua & @xmath2854 & @xmath2855 & @xmath2856 & @xmath2857 & @xmath2565 & & & + ps113eck & @xmath2858 & @xmath2859 & @xmath2860 & @xmath2861 & @xmath2862 & & & + ps113esn & @xmath2863 & @xmath2864 & @xmath2865 & @xmath2866 & @xmath2867 & & & | in recent years , wide - field sky surveys providing deep multi - band imaging have presented a new path for indirectly characterizing the progenitor populations of core - collapse supernovae ( sn ) : systematic light curve studies .
we assemble a set of 76 @xmath0-band type iip sn light curves from , obtained over a constant survey program of 4 years and classified using both spectroscopy and machine learning - based photometric techniques .
we develop and apply a new bayesian model for the full multi - band evolution of each light curve in the sample .
we find no evidence of a sub - population of fast - declining explosions ( historically referred to as `` type iil '' sne ) .
however , we identify a highly significant relation between the plateau phase decay rate and peak luminosity among our sne iip .
these results argue in favor of a single parameter , likely determined by initial stellar mass , predominantly controlling the explosions of red supergiants . this relation could also be applied for supernova cosmology , offering a standardizable candle good to an intrinsic scatter of @xmath1 mag .
we compare each light curve to physical models from hydrodynamic simulations to estimate progenitor initial masses and other properties of the type iip sn sample .
we show that correction of systematic discrepancies between modeled and observed sn iip light curve properties and an expanded grid of progenitor properties , are needed to enable robust progenitor inferences from multi - band light curve samples of this kind .
this work will serve as a pathfinder for photometric studies of core - collapse sne to be conducted through future wide field transient searches . |
the quasiclassical technique is one of the most powerful methods to tackle transport problems .
its main virtue relies in the fact that starting from a microscopic quantum formulation of the problem at hand it aims at deriving a simpler kinetic equation resembling the semiclassical boltzmann one . in deriving such an equation
some of the information at microscopic level is suitably incorporated in a set of parameters characterizing the physical system at macroscopic level .
since the first application to superconductivity , this equation is known as the eilenberger equation ( for a review see for instance @xcite ) .
we have recently derived@xcite such an equation for a two - dimensional electron gas in the presence of spin orbit coupling with hamiltonian @xmath1 where @xmath2 is a momentum dependent internal magnetic field . in the case of rashba spin - orbit coupling @xmath3 . in ref .
@xcite we adopted the standard @xmath0-integration procedure to arrive at the eilenberger equation , and , though this leads to correct results , we feel the need for a deeper understanding , which we provide in the present paper . in
so doing we follow an analysis carried out by shelankov@xcite .
finally , we use the eilenberger equation to study the response to an external electric field in the presence of magnetic impurities .
in deriving the eilenberger equation a key observation is that , by subtracting from the dyson equation its hermitian conjugate , one eliminates the singularity for equal space - time arguments and gets a simpler equation for the @xmath0-integrated green function @xmath4 here @xmath5 is the green function in wigner space , i.e. the fourier transform of @xmath6 with respect to the relative coordinate @xmath7 .
the `` check '' indicates that the green function is a @xmath8 by @xmath8 matrix in the keldysh space @xcite . to shed some light on the meaning of the @xmath0-integration ,
let us consider first the space dependence of the two - point retarded green function for free electrons in the absence of spin - orbit coupling @xmath9 at large distances , the integral is dominated by the extrema of the exponential under the condition of constant energy .
this forces the velocity to be parallel or antiparallel to the line connecting the two space arguments , @xmath10 .
it is then useful to consider the momentum components parallel ( @xmath11 ) and perpendicular ( @xmath12 ) to @xmath13 . given the presence of the pole , one can expand the energy in powers of the two momentum components @xmath14 . in the case of the retarded green function ,
the important region is that with velocity parallel to @xmath13 .
we then get @xmath15 one sees how the green function is factorized in a rapidly varying term @xmath16 , and a slow one , @xmath17 .
this suggests to write quite generally @xmath18 where @xmath19 is slowly varying and @xmath20 indicates the free green function .
explicitly , in the present equilibrium case @xmath21 for the advanced green function one can go through the same steps with the difference that the integral is dominated by the extremum corresponding to a velocity antiparallel to @xmath22 , so that one has the ingoing wave replacing the outgoing one . in the non - equilibrium case shelankov
has shown that @xmath23 and furthermore that the quasiclassical green function corresponds to the symmetrized expression @xmath24 when sending to zero the relative coordinate @xmath22 .
when the spin - orbit coupling is present the green function becomes a matrix in spin space and the fermi surface splits into two branches @xmath25 .
we always assume this splitting to be small compared to the fermi energy , i.e. @xmath26 . in the case of the rashba interaction
we write @xmath27 where @xmath28 is the projector relative to the @xmath29 energy branch and the curly brackets denote the anticommutator .
this ansatz allows us to proceed in wigner space as before , while retaining the information on the coupling and coherence of the two bands .
eq.([6bis ] ) is the equivalent in real space of the ansatz for the green function @xmath30 used in ref.@xcite .
with such an ansatz , eq.([6bis ] ) , we obtain from eq.([5special ] ) @xmath31 what we have explicitly shown for the retarded component of the green function can be extended to the advanced and keldysh components too .
notice that @xmath32 and @xmath33 coincide in the absence of spin - orbit coupling , since in that case @xmath34 .
the derivation of the eilenberger equation can now be done following the steps detailed in ref.@xcite .
we do not repeat them here and give just the final result @xmath35\big ) \nonumber \\ & = & - { \rm i } \left [ \check \sigma , \check g \right ] , \label{11}\end{aligned}\ ] ] where @xmath36 , @xmath37 and both the momentum @xmath38 and the internal field @xmath39 are evaluated at the @xmath40-branch of the fermi surface .
finally , @xmath41 is the self - energy .
it is often convenient to expand @xmath42 in terms of pauli matrices , @xmath43 , to explicitly separate charge and spin components .
physical quantities like charge and spin densities and currents are related to the keldysh component of @xmath42 .
for example the spin current for @xmath44 , @xmath45 is @xmath46 where @xmath47 and @xmath48 is the angle average over the directions of @xmath49 .
focusing on the rashba interaction , we study the effects of magnetic impurities on spin currents . in @xcite and @xcite
the problem has been recently tackled via diagrammatic techniques .
we show how analogous results can be obtained in a simple and rather elegant way relying on eq.([11 ] ) .
as it is well known , spin currents arising from the spin hall effect in such a system are completely suppressed by the presence of non - magnetic scatterers . by taking the angular average of eq.([11 ] ) , one obtains a set of continuity equations for the various spin components which let one easily understand the origin of this cancellation .
explicitly , by assuming @xmath50-wave and non - magnetic impurities randomly distributed in the system @xmath51 the self - energy in the born approximation turns out to be @xmath41 @xmath52 , @xmath53 being the momentum scattering rate .
the continuity equations for the @xmath45 spin components then read @xmath54 a rather important peculiarity of the rashba hamiltonian is that it lets one write the vector product appearing above in terms of the various spin currents , so that , by choosing for example @xmath55 , we are left with @xmath56 under stationary and homogeneous conditions this implies the vanishing of the @xmath57 spin current .
as soon as magnetic impurities are introduced in the system , their presence changes the self - energy and leads to the appearance of additional terms in eq.([12 ] ) .
we assume the magnetic scatterers to be also isotropic and randomly distributed @xmath58 and , proceeding again in the born approximation , we obtain the self - energy @xmath59 here @xmath60 is the spin - flip rate
. with this , and by considering again stationary and homogeneous conditions , eq.([12 ] ) becomes @xmath61 which in terms of the real spin current and polarization means @xmath62 by assuming a low concentration of magnetic impurities , we can use in eq.([fine1 ] ) the value of the @xmath63-spin polarization valid in their absence , @xmath64@xcite , @xmath65 being the external , homogeneous electric field .
we then get the spin hall conductivity to first order in @xmath66 @xmath67 a results that differs from those on refs.@xcite .
this is not surprising for ref.@xcite , which neglects normal impurity scattering and then considers the opposite limit .
the reason why our result does not agree with the low magnetic impurity - concentration limit of eq.(20 ) of ref.@xcite is not clear to us and deserves further investigation .
99 j. rammer and h. smith , rev .
phys . * 58 * , 323 ( 1986 ) .
r. raimondi , c. gorini , p. schwab , and m. dzierzawa , phys .
b * 74 * , 035340 ( 2006 ) .
a. i. shelankov , j. low temp .
phys . * 60 * , 29 ( 1985 ) .
j. inoue , t. kato , y. ishikawa , h. itoh , g. e. w. bauer , l. w. molenkamp , phys .
lett . * 97 * , 046604 ( 2006 ) .
p. wang , y. li , x. zhao , phys .
b * 75 * , 075326 ( 2007 ) .
v. m. edelstein , solid state commun .
* 73 * , 233 ( 1990 ) ; j. phys . : condens .
matter * 5 * , 2603 ( 1993 ) . | we discuss the quasiclassical green function method for a two - dimensional electron gas in the presence of spin - orbit coupling , with emphasis on the meaning of the @xmath0-integration procedure . as an application of our approach ,
we demonstrate how the spin - hall conductivity , in the presence of spin - flip scattering , can be easily obtained from the spin - density continuity equation .
ep2ds-17 , manuscript , latex-2e , style files 72.25.ba , 72.25.dc |
a subtitle to this talk should perhaps go like `` a see - saw - like mechanism without a majorana mass '' . here , i shall try to present arguments as to why it is interesting and worthwhile to study scenarios in which neutrinos possess a mass which is _ pure dirac _ in nature . along the way , i shall try to argue that one should perhaps try to separate the issue of a see - saw like mechanism from that of a majorana mass . by see - saw - like mechanism
, it is meant that a `` tiny '' mass arises due to the presence of a very large scale .
the suggestions that neutrinos do indeed possess a mass came from three different sources , all of which involve oscillations of one type of neutrino into another type .
they are the superkamiokande atmospheric neutrino oscillation , the solar neutrino results , and the lsnd result @xcite . the present status of these three oscillation experiments is well presented in this workshop .
the future confirmation of all three will certainly have a profound impact on the understanding of the origin of neutrino masses .
in particular , it is now generally agreed that if there were only three light , active ( i.e. electroweak non - singlet ) neutrinos , one would not be able to explain all three oscillation phenomena .
the confirmation of _ all three results _
would most likeky involve the presence of a sterile neutrino .
whatever the future experiments might indicate , one thing is probably true : if neutrinos do have a mass , it is certainly tiny compared with all known fermion masses .
typically , @xmath0(electroweak scale ) .
why is it so small ?
is it a dirac or a majorana mass ?
this last question presently has no answer from any known experiment .
the nature of the mass will no doubt have very important physical implications .
the route to a gauge unification will certainly be very different in the two cases . whether or not the mass is dirac or majorana , there is probably some new physics which is responsible for making it so tiny .
what is the scale of this new physics ?
what are the possible mechanisms which could give rise to the tiny mass ? in trying to answer these questions , one can not help but realize that there is something _ very special _ about neutrinos ( specifically the right - handed ones ) which make them different from all other known fermions .
do they carry some special symmetry ?
one example of new physics which might be responsible for a small neutrino mass is the ever - popular and beautiful see - saw mechanism of gell - mann , ramond and slansky @xcite , in which a majorana mass arises through a lepton number violating process .
generically , one would have @xmath1 , with @xmath2 electroweak scale , and @xmath3 some typical gut scale .
since one expects @xmath4 , one automatically obtains a tiny _ majorana _ neutrino mass .
the actual detail of the neutrino mass matrix is however quite involved and usually depends on some kind of ansatz .
but that is the same old story with any fermion mass problem anyway .
the crucial point is the fact that the very smallness of the neutrino mass comes from the presumed existence of a very large scale @xmath5 compared with the electroweak scale .
this mechanism has practically become a standard one for generating neutrino mass . why then does one bother to look for an alternative ?
first of all , there is so far _ no evidence _ that , if neutrinos do have a mass , it should be of a majorana type . if anything , the present absence of neutrinoless double beta decay might indicate the contrary .
( strictly speaking , what it does is to set an upper limit on a majorana mass of approximately 0.2 ev , although actually it is a bound on @xmath6 .
therefore , this question is entirely open . in the meantime
, it is appropriate and important to consider scenarios in which neutrinos are pure dirac .
the questions are : how can one construct a tiny _ dirac _ mass for the neutrinos ?
how natural can it be ?
can one learn something new ?
are there consequences that can be tested ?
the construction of the model reported in this talk was based on two papers @xcite .
there exists several other works @xcite on dirac neutrino masses which are very different from @xcite .
the first one @xcite laid the foundation of the model .
the second one @xcite is a vastly improved and much more detailed version , with new results not reported in @xcite . in constructing this model , we followed the following self - imposed requirements : \1 ) the smallness of the dirac neutrino mass should arise in a more or less natural way .
\2 ) the model should have testable phenomenological consequences , other than just merely reproducing the neutrino mass pattern for the oscillation data .
\3 ) one should ask oneself if one can learn , from the construction of the model , something more than just neutrino masses .
this also means that one should go beyond the neutrino sector to include the charged lepton and the quark sectors as well .
this last sentence refers to work in progress and will not be reported here . before describing our model ,
let us briefly mention a few facts .
first of all , it is rather easy to obtain a dirac mass for the neutrino by simply adding a right - handed neutrino to the standard model .
this right - handed neutrino ( one for each generation ) is an electroweak singlet and , as a result , can have a gauge - invariant yukawa coupling : @xmath7 . the dirac neutrino mass would then be @xmath8 . with @xmath9 , a neutrino mass of o(1 ev ) would require a yukawa coupling @xmath10 .
although there is nothing wrong with it , a coupling of that magnitude is normally considered to be extremely fine - tuned , if it is put in by hand !
could @xmath10 be _ dynamical _ ?
would the limit @xmath11 lead to some new symmetry ? what would it be ?
this new symmetry would be the one that protects the neutrino mass from being `` large '' . in choosing such a symmetry , we followed our self - imposed requirement # 3 : one should learn something more from it than just merely providing a symmetry to protect the neutrino mass .
first , in order to implement the symmetry protection , one should assume that this new symmetry is particular to the neutrinos , in particular the right - handed ones since left - handed neutrinos are weak interaction partners of standard charged leptons .
therefore , it will be assumed that all fermions other @xmath12 s are _ singlets _ under this new symmetry .
one of the reasons we adhere to our requirement # 3 is the wish to work from the bottom up , instead of from the top down . as a result
, we would try to make every increased step in energy as meaningful as possible .
the symmetry chosen in @xcite is a chiral gauge symmetry .
it is @xmath13 , where the subscript @xmath12 means that _ only _ @xmath12 s carry @xmath13 quantum numbers .
why @xmath13 ? because it is a chiral gauge @xmath14 which has a very important property : for weyl fermions transforming as doublets under such a group , there exists an argument due to witten @xcite that says , in a nutshell , that , because of the presence of a non - perturbative global anomaly the number of such weyl doublets has to be _ even _ in order for the theory to be well - defined .
( in the language of quantum field theory , this means that the generating functional should be non - vanishing . )
amusingly enough , a long - forgotten fact about the sm is related to the witten anomaly .
the absence of such anomaly for @xmath15 require an even number of electroweak doublets per family , which is the case there : one lepton and three quark doublets .
( from a historical prespective , one might say that , had this constraint be known in the early seventies , the sm , as we now know it , would have had an extra strong argument in its favor , prior to the slac experiment . ) in our case , let @xmath12 transform as doublets under @xmath13 , i.e. we now have @xmath16 .
( cosmological issues concerning @xmath17 s are discussed in @xcite . )
the absence of the witten anomaly then requires the number of @xmath18 s to be _
even_. if furthermore , @xmath18 s carry some family indices ( if a family symmetry exists ) then this constraint can have a profound implication on the issue of family replication .
further remarks can be found in @xcite .
we know that families _ do mix_. in consequence , we need some kind of family symmetry .
the family symmetry chosen in @xcite is a gauge symmetry .
this choice is pure prejudice : we believe that gauge theories are better choices for a family symmetry because of the fact that they do provide strong constraints on matter representations and because one might want to mimic the vertical symmetry ( the electroweak interactions ) .
we choose @xmath19 as our family gauge group with all fermions transforming as _ vector _ representions in order to avoid the usual traingle anomaly .
our model is given by the following extension of the sm : @xmath20 in @xcite , i have discussed the various arguments used to constrain @xmath21 . in this talk ,
i shall however restrict @xmath21 to be @xmath22 .
this means that this is a four - family model .
is a 4th generation ruled out by experiment as one often hears ?
the answer is : not at all ! for instance , the usual question is the following : what about the z width which tells us that there are only three light neutrinos ?
this does not apply to the case when the 4th neutrino is more massive than half the z mass
. why then would it be so heavy when the other three neutrinos are so light ?
is nt it unnatural ?
the answer is no as we shall see below .
then , what about the 4th generation quarks and charged leptons ? there exists a review @xcite dealing extensively with this question .
a quick summary of that review is the statement that there is plenty of room for the discovery of the 4th generation , either at the next upgraded collider experiments at the tevatron , or at the lhc .
i shall now turn to the basic results of the model .
i shall describe the results which are based solely on the assumption that only two oscillation results are correct : the solar and atmospheric neutrino oscillation data .
i shall mention at the end a possibility in case all three oscillation experiments are confirmed . in a nutshell , here are the results obtained in @xcite .
1 ) we obtain three light , _ near degenerate _ neutrinos .
2 ) the `` tiny '' masses are obtained _ dynamically _ at one loop
. one will see below the reason for the use of the term `` see - saw - like mechanism with dirac mass '' .
3 ) the masses of the light neutrinos , @xmath23 ( @xmath24 ) , and @xmath25 are correlated in an interesting way : a ) if the msw solution is chosen for the solar neutrino problem , the masses can be as large as o(few ev s ) and can provide enough mass for the hot dark matter ( hdm ) ; b ) if the vacuum solution is chosen instead , the masses are found to be at most @xmath26 0.1 ev and , as a result , are too small to be relevant to the hdm .
4 ) there are a number of phenomenological consequences which can be tested : there is _ no _ neutrinoless double beta decay since the mass is dirac ; there is a possibility of detection of `` light '' ( a couple of hundreds of gev s ) vector - like fermions ; etc ... 5 ) there are a number of possible cosmological consequences : baryon asymmetry through neutrinogenesis with a pure dirac neutrino mass ; perhaps some of the very heavy vector - like fermions could be the source of ultra high energy cosmic rays . in writing down the lagrangian for our model ,
we take into account the fact that our point here is to obtain a pure dirac mass . therefore ,
b - l will be assumed .
the particle content is listed in the table .particle content and quantum numbers of @xmath27 [ cols= " < , > " , ] and the lagrangian is given by @xmath28 after integrating out the @xmath29 , @xmath30 , and @xmath31 fields , the relevant part of the effective lagrangian below @xmath32 reads @xmath33 where @xmath34 as one can see , all neutrinos are _ massless _ when @xmath35 is unbroken .
assume @xmath36 and @xmath37 , the 4th neutrino gets a mass @xmath38 with @xmath39 .
one can arrange the masses and couplings in such a way that @xmath40 o(1 ) , and @xmath41 o(100 gev ) .
there is nothing unnatural about such a choice .
it is _ natural _ in this scenario to have the 4th neutrino having a mass of o(100 gev ) .
one point which is worth emphasizing again is the following : the breaking of @xmath13 ( in addition to @xmath42 ) is essential for @xmath43 to be non - vanishing ! at this stage ( tree - level ) , there are _ three massless _ neutrinos .
it turns out , at one loop , that the three formely - massless neutrinos acquire a common mass , i.e. they are _ degenerate _ : @xmath44 where @xmath45 are masses of particles which participate in the loop diagram and where @xmath46 .
this is shown in fig . 1 . here
@xmath47 is given by @xmath48}{(m_g^2-m_f^2)(m_p^2-m_f^2 ) } - \nonumber \\ & & ( m_f \leftrightarrow m_2)\}. \label{int1}\end{aligned}\ ] ] one important remark is in order here . from eq .
( [ rat1 ] ) , one notices that the neutrino mass does not depend explicitely on the value of the masses @xmath45 but only on their _
ratios_. if one takes @xmath49 as a `` base '' mass for example , it turns out that one can obtain quite a small mass for the light neutrinos , @xmath50 , as long as one has , e.g. , @xmath51 , with @xmath49 being an arbitrary number which can be as low as experimentally allowed i.e. o(200 gev ) @xcite .
( remember that @xmath29 stands for @xmath52 , where @xmath53 and @xmath54 are degenerate in mass . )
this can be seen from fig .
2 . just to illustrate this point , a numerical example would be helpful .
take @xmath55 gev , @xmath56 , @xmath57 , @xmath58 , we obtain @xmath59 ev .
i wish to emphasize that this is _ not _ a prediction .
it will have to come from some deeper theory which will fix the above mass ratios and hence @xmath60 .
this is only meant to illustrate the fact that , in our model , it is quite natural to get a teeny , tiny dirac neutrino mass . to go further
, one needs to lift the degeneracy of the light neutrinos . before showing how one would go about doing it ,
let us explain what we meant by `` see - saw - like mechanism without a majorana mass '' .
the function @xmath61 shown in eq .
( [ int1 ] ) has the following limit : @xmath62 for @xmath63 . from eq .
( [ rat1 ] ) , one can see that , in the above limit , @xmath64 .
if @xmath65 ( @xmath66 ) , one would obtain @xmath67 this is a typical see - saw - like relation ! with a `` low '' mass ( @xmath26 electroweak scale ) vector - like fermion , @xmath29 , one can qualitatively see that @xmath68 can be _ very small _
when @xmath69 .
this behaviour is very reminescent of the see - saw mechanism , except that , in our case , the mass is _
dirac_. now the next step is to introduce mixing among the neutrinos in order to lift the mass degeneracy .
this can be acomplished , in our model , by involving mixing in the scalar sector . at this stage
, the degeneracy among the three light neutrinos is due to a remaining global @xmath70 symmetry .
this remaining global symmetry can be explicitely broken by the scalar sector as shown in @xcite .
the offshoot of all this is , in the end , the fact that this explicit breaking depends on a parameter denoted by @xmath71 in @xcite .
it turns out that , in a paper under preparation @xcite , @xmath71 itself will be severely constrained when our model is extended to the quark sector .
this is because the same scalar sector is also involved in the quarks .
it is satisfying to see the link between the quark and lepton sectors .
this is however not the subject of this talk and i shall now return to the task at hand .
the first case which was investigated in @xcite is when the scalar sector is written down in such a way that there is no mixing between the 4th neutrino and the other three in the mass matrix .
something interesting happens here .
it turns out that the mass splittings are quasi - degenerate , in the sense that @xmath72 . if this model were to explain both solar and atmospheric oscillation data , this quasi - degeneracy of @xmath25 has to be lifted . also , the fact that the oscillation data appear to show @xmath73 implies , in the context of our model , that indirectly the data suggests the existence of a 4th neutrino whose mixing with the lighter three will lift the quasi - degeneracy of @xmath25 . before showing how this could be done ,
let us see what these results imply .
if the vacuum solution for solar neutrinos is preferred , i.e. @xmath74 , then it is found that the median mass value of three almost degenerate neutrinos is @xmath75 .
as stated earlier , this is not enough for the hdm scenario .
if the msw solution is preferred , i.e. @xmath76 , the median mass value could be @xmath77 o(few ev s ) , a reasonable value for the hdm scenario .
the fact that @xmath73 indicates , in the context of this model , the existence of more than three light neutrinos . in @xcite where only the atmospheric and solar data were taken into account
, this means that it indicates the existence of a 4th neutrino .
again through the scalar sector , one can construct e.g. a mixing between the 3rd and 4th neutrino ( other possibilities exist ) .
the size of the mixing determines the correct mass splittings .
it was found that there are strong constraints on the some of the scalar masses when one requires that @xmath78 and @xmath79 .
the next question concerns the oscillation angles . to find out what they are
, one needs to know the leptonic `` ckm '' matrix : @xmath80 . in dealing with the neutrino sector of our model ,
we have presented a case where @xmath81 can be computed .
it is basically given by : @xmath82 as for @xmath83 which requires a detailed study of the charged lepton sector , a construction is in progress .
in the meantime , just for the purpose of illustration , reference @xcite has been used in which a simple ansatz for the charged lepton sector was given .
the reason for using this reference is because it contains an ansatz for @xmath81 which is similar to ours .
therefore the results should be similar : a small angle msw solution and a large angle atmospheric solution .
i have presented in this talk a model which can `` naturally '' give rise to a teeny , tiny dirac neutrino mass , without resorting to the concept of a majorana mass .
what was shown was the need to differentiate between the see - saw mechanism and the existence of a majorana mass . in this model ,
the smallness of the light neutrino mass arises in a see - saw - like fashion , with the mass being purely dirac .
as we have argued in the introduction , the reason for constructing such a model is twofold : a ) one does not know experimentally whether the mass is majorana or dirac ; b ) the physics is _ very _ different in the two cases . at this stage of our knowledge , it is perhaps prudent to explore all different possibilities .
since so much has already been worked out with models using a majorana mass , any new model which takes a different route should have a clear motivation and predictable consequences . for our model , we have presented clearly our motivation : naturally small dirac neutrino mass , family replication , etc .. ; and predictions concerning the neutrino sector : vacuum solution @xmath84 hdm , msw solution @xmath85 hdm , @xmath86 as an indirect indication of a 4th heavy neutrino .
other phenomenological implications include : \2 ) the existence of `` long - lived '' and `` light '' ( i.e. @xmath87 ) vector - like leptons ( @xmath29 ) whose detection might be possible at the lhc .
a study of this kind of search can be found in a comprehensive review @xcite .
the quark counterparts should also be detectable @xcite .
there are several other phenomelogical issues to be discussed . for lack of space , a few of those
will be briefly mentioned .
one is the s parameter for example .
it is well - known that , to leading order , vector - like fermions which carry electroweak @xmath15 quatum numbers _ do not _ contribute to s if one has a _ degenerate _
@xmath14 doublet .
the reason for this being so is because the right - handed contribution cancels exactly the left - handed contribution .
therefore , to leading order , there is no constraint from the s parameter on the mass of the @xmath29-fermions .
this point and other issues concerning quarks and leptons beyond the third generation are discussed in @xcite .
issues such as the decay of the heavy 4th neutrino can be consulted in @xcite . also , another issue such as the magnitude of flavor - changing neutral currents , e.g. @xmath88 , will be discussed in an upcoming paper dealing with the charged lepton sector . however , a preliminary statement can be made .
for example , in the case of @xmath88 , there are two kinds of contributions : one coming from the propagation of neutrinos with a non - zero mass inside the loop diagram for the process , and the other one coming from diagrams involving the new vector - like fermions .
it turns out that both contributions are negligible : 1 ) in the first case , it is because @xmath89 ; 2 ) in the second case , it is because of the cancellations of the type described in @xcite .
\1 ) can the fermion @xmath31 be the source of ultra high energy cosmics rays ( e @xmath90 ) ?
for example if @xmath91 then @xmath92 ?
would the decays of @xmath31 ( e.g. @xmath93 high energy quarks and leptons ) be responsible for uhecr ?
this deserves a closer look .
\2 ) the possibility of baryon asymmetry from neutrinogenesis .
this is a scenario of ref .
the ingredients needed for such a scenario to work are basically : 1 ) a tiny , pure dirac neutrino mass ; 2 ) a decay process from some superheavy particles at the gut ( or similar ) scale into right - handed neutrinos such that there is an asymmetry between right - handed neutrinos and anti - neutrinos ; 3 ) a b+l violating process from the electroweak sphaleron .
since the dirac neutrinos have a tiny yukawa coupling ( which is dynamical in our case ) , the part of b+l which is stored in the right - handed neutrinos , @xmath94 , survived the sphaleron `` washout '' .
so if one starts out with b - l=0 , this process can generate a _ net _
baryon number , @xmath95 .
one last remark is in order here .
if all three oscillation experiments were to be confirmed in the future , there seems to be a need for a sterile neutrino .
how will it fit in our framework ?
it turns out that some modifications of the previous analysis will be needed but the basic framework is still the same .
this work is in progress .
e - mail : [email protected] .
see peter fisher , boris kayser , and kevin s. mcfarland , hep- ph/9906244 , for a comprehensive review and a list of references .
m. gell - mann , p. ramond , and r. slansky in supergravity , edited by d. freedman et al .
( 1979 ) ; t. yanagida , kek lectures ( unpublished ) ( 1979 ) ; r. n. mohapatra and g. senjanovc , phys . rev
44 * , 912 ( 1980 ) .
p. q. hung , phys .
rev . * d59 * , 113008 ( 1999 ) .
p. q. hung , phys .
* d * ( in press ) ( september , 2000 ) , hep - ph/0003303 .
b. balakrishna and r. n. mohapatra , phys . lett . *
b216 * , 349 ( 1989 ) ; d. chang and r. n. mohapatra , phys .
58 * , 1600 ( 1987 ) ; g. barenboim and f. scheck , phys . lett . *
b461 * , 1235 ( 1999 ) ; j. i. silva - marcos , phys . rev .
* d59 * , 091301 ( 1999 ) .
e. witten , phys .
b117 * , 324 ( 1982 ) .
p. h. frampton , p. q. hung , and m. sher , physics reports * 330 * , 263 ( 2000 ) .
p. q. hung and a. soddu , manuscript under preparation ( 2000 ) .
m. fukugita , m. tanimoto , t. yanagida , phys . rev . *
d57 * , 4429 ( 1998 ) .
k. dick , m. lindner , m. ratz , and d. wright , hep - ph/9907562 v3 . | there are now strong hints suggesting that neutrinos do have a mass after all .
if they do have a mass , it would have to be tiny .
why is it so ?
is it dirac or majorana ?
can one build a model in which a teeny , tiny dirac neutrino mass arises in a natural way ?
can one learn something else other than just neutrino masses ?
what are the extra phenomenological consequences of such a model ? these are the questions that i will try to focus on in this talk . |
neutron star low mass x - ray binaries ( lmxb ) were believed to behavior similarly to the black hole lmxbs ( lin et al . 2007 ) .
they spend most of the time in quiescence , and occasionally show an outburst with accretion rate dramatically increased . during the outburst
, the neutron star lmxbs will evolve at the similar track of the state transition as black hole lmxbs do , basically from a low / hard state to a high / soft state ( remillard & mcclintock 2006 ) .
the optical emission of lmxbs was often contributed to thermal emission of the companion star and outer accretion disk , and sometimes synchrotron emission of a jet .
the disk can be heated by mainly two physical processes , one is the friction between adjacent layers of the disk when the materials were accreted ( shakura & sunyaev 1973 ) , the other is the x - ray irradiation from the inner disk ( van paradijs & mcclintock 1994 ) . with the evolution of disk structure and other physical properties , the radiation mechanism of the optical flux will vary , which will be different for neutron star lmxbs and black hole lmxbs . for black hole lmxbs ,
the synchrotron emission of a jet was believed to dominate the optical emission during low / hard state , with possible disk - reprocessed emission ( russell et al .
2006 ) . in the soft state ,
all the near - infrared and some of the optical emissions are suppressed , a behavior indicative of the jet switching off in transition to the soft state ( russell et al .
the viscously heated disk emission may become the dominant source .
the multi - wavelength observation of gx 339 - 4 provided a perfect example to demonstrate the trend : a good correlation among the fluxes from the x - ray power - law component , the radio , and the optical was found during low / hard state which suggests the jet emission , however , the optical flux dramatically decreased when the source entered high / soft state .
meanwhile , an @xmath3 two - week x - ray flux delay was found during high / soft state which indicates a viscously heated disk emission . for neutron star lmxbs ,
the jet emission is not important unless at very high luminosity .
x - ray reprocessing was believed to dominate the optical emission at low / hard state , with possible contribution from viscously heated disk ( russell et al .
the long - time observation of neutron star lmxb , e.g. aquila x-1 hr ) and therefore a very large disk .
usually transient nss have orbital periods in the 2 - 8 hr range .
however , the quasi - simultaneous optical and x - ray luminosity of aquila x-1 follows the general relation for neutron star lmxbs ( russell et al .
2006 ) ] , shows that neither the optical / near - infrared color nor its brightness change sharply during an x - ray spectral state transition .
so it is believed that for aquila x-1 the outer accretion disk is not affected by x - ray spectral state transitions ( maitra & bailyn 2008 ) , the x - ray reprocessing was thought to contribute most of optical emission at both the low / hard and high / soft state .
when the optical emission is dominated by the viscously - heated - disk emission , the emission at each radius provides a measure of the instantaneous local accretion rate at the given radius .
the x - ray and optical emission , respectively , map the mass flow through the inner and outer disk
. continuous monitoring to both the x - ray and optical emission allows us to track the temporal evolution of the system .
the cross - correlation of the x - ray and optical light curves helps to map the accretion flow direction , the x - ray / optical time delay reflects the viscous timescale of the disk ( e.g. homan et al . 2005 ) .
the time lag between the initial point of the outburst in x - ray and optical emission was believed to be able to , to some extent , test the disk model and the trigger mechanism of the outburst ( narayan et al .
1996 ) .
sax j1810.8 - 2609 is a soft x - ray transient ( sxt ) discovered on 10 march 1998 with the wide field cameras ( 2 - 28 kev ) onboard the _ bepposax _ satellite ( ubertini et al . 1998 ) .
it was identified as a neutron star lmxb because a strong type - i x - ray burst was detected ( natalucci et al .
the distance was estimated to be @xmath34.9 kpc . on 11 - 12 march 1998 , a follow - up target of opportunity ( too ) observation with the narrow field instrument onboard _
bepposax _ was performed with a total observing time of 85.1 ks .
it showed a hard x - ray spectrum with emission up to 200 kev . the broadband spectrum ( 0.1 - 200 kev )
can be described by two components : a soft black body component with the temperature @xmath4 0.5 kev , and a power - law component with the photon index @xmath5 ( natalucci et al . 2000 ) . from 1998 through 2007 ,
sax j1810.8 - 2609 had been in a quiescent state .
the neutron star system in quiescence was also detected by _
chandra _ on 16th august 2003 ( jonker et al . 2004 ) .
it had an unabsorbed x - ray luminosity of @xmath310@xmath6 erg s@xmath7 over the energy range of 0.3 - 10 kev , given the distance of 4.9 kpc .
it shows that the quiescent spectrum could be well fitted by the two models : neutron star atmosphere + power - law model and black body + power - law model . in august 2007 , _
swift _ detected a new phase of highly luminous activity ( parson et al .
2007 ) , and the luminosity varies between ( 1.1 - 2.6 ) @xmath8 erg s@xmath7 during this outburst . considering the time interval of the recurrence , the observed outburst luminosity corresponds to a low time - averaged accretion rate of @xmath9 @xmath10 @xmath11 ( fiocchi et al .
the x - ray spectra had shown the evolution during different epochs of the outburst , but a significant power - law component was always present ( fiocchi et al .
it is noted that the source never reached the high soft state during the outburst . in this paper , we obtained the twelve - year optical light curve of sax j1810.8 - 2609 , which covers two outbursts .
we cross - correlated the optical light curve with the x - ray light curve from the _ rxte_/asm and _ swift_/bat archive .
section 2 describes the observations and data calibration . in section 3
we identify the optical counterpart and analyze the temporal morphology of the outburst . in section 4
we show the results of the cross correlation and discuss their implications .
section 5 is the summary .
the region containing sax j1810.8 - 2609 has been regularly observed by the optical gravitational lensing experiment ( ogle ) project during eleven observing seasons from 1997 to 2009 , including the outbursts in 1998 march and 2007 august .
the ogle data were collected using a dedicated 1.3 m warsaw telescope located at las campanas observatory , chile , and operated by the carnegie institution of washington . during the years of 1996 - 2000
ogle used a single 2048@xmath122048 ccd chip ( with a pixed size of 0.417 ) operating in the drift - scan mode ( ogle - ii ; udalski et al .
1997 ) . for ogle - iii ( 2001 - 2009 )
the camera was upgraded to a mosaic of eight @xmath13 ccd chips with pixel size of 0.26(udalski 2003 ) . during ogle - ii and ogle - iii , for the bulge fields
each observation was exposed for 120 seconds .
the observations were usually done in the i - band .
datasets from both phases of ogle were processed using the same photometric package based on difference image analysis ( dia ; wozniak 2000 ) .
the method requires creation of a template / reference image , composed of the best available images of a given field . the template is then convolved with the psf of an image to be analyzed and the subtraction is performed , revealing residuals due to changes in the objects fluxes , or to the appearance of new objects .
the flux residuals are then measured with aperture photometry , and light curves are created ( udalski 2008 ) . in templates for both ogle - ii and ogle - iii we found a faint star with a constant @xmath14 mag , 0.44away from the outburst position as shown in figure 2 . because there was no object at the outburst position on the templates , all the residual flux from the outbursts
was attributed to this faint star .
therefore the composite light curve , shown in figure 1 , contains both the baseline fluxes from the template object and outbursts , and the faint template object contributes a negligible amount of emission during outbursts .
this source is labeled in the microlensing observations in astrophysics ( moa ) microlensing alert catalogue as moa_2007_bul_365(bond et al .
moa was also operating in the i - band . in quiescence , this source is also well below the detection threshold of moa .
it was monitored during the 2007 outburst .
we calibrated the moa light curve using the quasi - simultaneous ogle light curve .
the all sky monitor ( asm ) ( 2 - 12 kev ) onboard the rossi x - ray timing explorer ( _ rxte _ ) has provided the longest continuous coverage of luminous x - ray sources ( levine et al .
in its quiescent state , sax j1810.8 - 2609 was not detected by the asm , but the asm detected and monitored both outbursts .
the second outburst was also monitored by _ swift _ burst alert telescope ( bat ; @xmath315 - 50 kev ) through the hard x - ray transient monitor program ( barthelmy et al . 2005 ) .
we retrieved both data sets from their respective websites .
in addition , _
integral_/ibis ( 22 - 68 kev ) also detected this source during this outburst ( fiocchi et al .
2009 ) , however , we did not include the _ integral _ data because the monitoring was not frequent enough near peak .
in the optical band , the outburst is located in ogle iii field blg251.8 with the coordinates ( r.a=18@xmath1510@[email protected] , dec.=-26@xmath1809@[email protected] ) .
this position is consistent with the position of sax j1810.8 - 2609 ( r.a.=18@xmath1510@[email protected] , dec.=-26@xmath1809@[email protected] ) determined by _
( jonker et al 2004 ) , with an uncertainty of 0.6 .
it is also consistent with the optical counterpart position of ( r.a.=18@xmath1510@[email protected] dec.=-26@xmath1809@xmath1900@xmath20 ) with an uncertainty of 1determined by greiner et al ( 1998 ) .
the non - detection by moa and ogle constrains the source to be fainter than 21 mag in i band during quiescence ( figure 2 ) . in the optical light curve ,
we indicate the 1998 outburst with light blue ` @xmath23 ' and the 2007 outburst with dark blue ` @xmath24 ' ( see figure 1 ) . it is noted that the light blue points are concentrated on the decay phase in 1998 outburst because the main part was not observed .
the optical observations were usually done in the i - band , however , one observation in r band were also available .
we can use the magnitude difference between i and r bands to estimate the disk temperature in the outer region .
the r - band observation is r=@xmath25 on march 13th 1998 ( greiner et al .
1998 ) , and two ogle i - band observations closest in time are : i=@xmath26 mag on march 10 and i=@xmath27 mag on march 16 , respectively .
so we take the average for the two i - band detections , and use it as the proxy for the i - band magnitude on march 13th .
therefore , we have r - i@xmath28 mag . once we make the reduction correction ( sumi 2004 ) , it is r - i@xmath29 mag , and it corresponds to a temperature @xmath30 k with a black body model , which is consistent with the typical temperature for the outer disk .
the 2007 august outburst was well monitored , and the whole optical and x - ray light curves are shown in figure 3 . the hard ( 15 - 50 kev ) and soft ( 2 - 10 kev ) x - ray data are from _ swift_/bat and _ rxte_/asm , respectively .
the moa and ogle data were combined together to form the optical light curve .
the similarity of the optical and the x - ray light curve during 2007 outburst confirms the identification of the optical counterpart .
we did not use the _ integral_/ibis ( 22 - 68 kev ) light curve because of the incomplete covering . in figure 1
, we notice that the fluxes at four occasions ( which are all out of the outburst phase ) were significantly higher than the baseline .
it is certain that three of these are not likely to be real because the other stars in the field also exhibited the similar flux feature during these observations .
it means that these flux measurements were affected by the systematic uncertainties that may be related to weather , sky background , or instrumental effects .
the fourth point with an elevated flux at mjd = 52583.5 is indicated with a red circle .
we do nt have any direct evidence to show this point is also affected , but the quality of the image makes it difficult for us to conclude that the system was experiencing a true increase in optical flux . if the flux increase is true , this data point is pretty interesting , and it may represent an short - duration ( no longer than 2 days ) outburst .
we checked all the dwell data of the corresponding _
rxte_/asm observations and found no x - ray detection .
this single flux increase may imply a magnetic loop reconnection in the outer disk as proposed by zurita et al ( 2003 ) .
the 2007 august outburst showed a multi - peak morphology ( see figure 3 ) .
we divided the whole outburst light curve into three parts : ( 1 ) an initial rise ( @xmath31 mjd @xmath32 ) , where the outburst is detected in optical but not in x - ray ; ( 2 ) the first peak ( @xmath33 mjd @xmath34 ) ; and ( 3 ) the rebrightening stage ( @xmath35 mjd @xmath36 ) , which includes all the subsequent peaks . the temporal behavior in the optical and x - ray bands is very similar .
the solid lines in figure 3 is used to separate these three different stages .
the position of significant optical peaks / dips were indicated with dotted lines .
the corresponding x - ray peaks / dips usually follow the optical peaks / dips but with a - few - days lag .
the first optical peak occurred at @xmath37 , while the corresponding x - ray flux peaked slightly later .
the optical light curve has a significant dip when it tends to get a second peak at @xmath38 , then a sudden rise roughly three days later .
the dip was followed by the x - ray light curve about two days later .
the third significant optical peak occurred at @xmath39 , and was followed by both the soft and hard x - ray light curve a few days later .
the optical light curve also has a wide peak near @xmath40 which is very similar to that in the hard x - ray light curve ; the optical emission still leads the x - ray emission .
the source was not detected by _
rxte_/asm until the flux reached the level of @xmath41 mcrab on mjd 54323 ( levine et al .
1996 ) , but the optical emission was detected as early as on mjd @xmath42 . during the initial rise phase ,
the optical detection ( ogle and moa ) was earlier than the x - ray detection ( _ rxte_/asm ) by @xmath43 days .
similar x - ray delays were found in quasi - simultaneous x - ray and optical monitoring of sxt gro j1655 - 40 ( orosz et al . 1997 ) , xte j1550 - 564 ( jain et al . 2001 ) , 4u 1543 - 47 ( buxton & bailyn 2004 ) and aquila x-1 ( shahbaz et al . 1998 ; maitra & bailyn 2008 ) , among which the initial point of outburst in x - rays lagged in optical by @xmath44 days .
the x - ray delays at the initial rise reflect the timescale of the thin disk , which determines the time it takes for the accretion flow to arrive at the inner region where the x - ray emission comes from .
it already takes into account the timescale by the material to fill in the _ advection - dominated accretion flow _ ( adaf ) region , which can extend to large radius in quiescent state ( e.g. narayan & yi 1995 ) .
for all the systems aforementioned , the initial x - ray rises of the outbursts were all determined by linear extrapolations of the _ rxte_/asm light curve .
we notice , however , that the detection threshold for _ rxte_/asm is much higher than the flux of the source in quiescence state .
the x - ray flux might have increased by a factor of a few tens before detected .
furthermore , the flux may actually increased in a way much slower than linear increase before detected by _
rxte_/asm , the initial point found by linear extrapolation may be actually much later than the actual starting point ( homan et al . 2005 ) . here
we apply a power - law extrapolation to the x - ray / optical data points during the rise phase of the outburst as shown in figure 4 .
it would make a big difference if the flux increased exponentially rather than linearly .
the x - ray luminosity was calculated with a distance of 4.9 kpc ( natalucci et al .
2000 ) . based on _ chandra _ observations ,
its quiescent luminosity is @xmath45 erg s@xmath7 ( jonker et al .
figure 4 shows an extrapolation that indicates that optical flux may have begun to rise at mjd @xmath46 with @xmath47 magnitude in quiescence .
the fit to the x - ray data has large uncertainty , it gave an x - ray initial rise at at mjd @xmath48 .
the start points of optical and x - ray flux are consistent with each other within one - sigma error .
it is significantly different from the start points directly detected .
we can not conclude x - rays delay at the start of the outburst due to the limited data .
we conducted cross - correlations of the optical and x - ray emission for the first peak and the rebrightening phases , respectively . during the first peak ,
we only cross - correlated the soft x - ray ( _ rxte _ , 2 - 12 kev ) and optical light curves .
however , during the rebrightening stage , we cross - correlated optical and soft x - ray , optical and hard x - ray ( _ swift _ , 15 - 50 kev ) , soft and hard x - ray emission , respectively . for the convenience of performing the cross - correlation , we did a linear interpolation to each light curve with a time interval of 0.1 days .
the cross - correlation between two interpolated light curves gave out a _
cross correlation factor _ ( ccf ) function with uniform time lags .
the factor ccf will tell us how strong the correlation is , and it is defined as @xmath49 , where @xmath50 and @xmath51 represent the data series with finite length , @xmath52 and @xmath53 are the averages of @xmath50 and @xmath51 , @xmath54 and @xmath55 are the variance of @xmath56 and @xmath57 , respectively .
one caveat is that the value of ccf may not be zero even if the two data series are not correlated at all , e.g. , if the data series are short , the noise may produce a non - zero ccf .
we therefore conducted a monte carlo ( mc ) simulation to create faked x - ray and optical light curves using the same temporal interval and uncertainty of each data point from the observed ones . in order to make sure
the faked light curves have no correlation , we assume each value on the faked light curves is taken from a gaussian distribution ( its central value is set zero , and @xmath58 of the distribution is set to be same as the observed uncertainty at each epoch ) .
therefore , the faked light curves generated in this way are like the pure noise .
once one set of faked light curves are generated , we did a linear interpolation and cross - correlated the interpolated light curves , obtaining the time lag and the maximum ccf . by repeating this process for 2000 times for the first peak phase and the rebrightening phase , respectively , we computed the distribution of the maximum values of ccf , and then derived the @xmath59 confidence level of ccf of each phase .
the @xmath59 confidence level for the cross - correlation for the first peak is ccf@xmath60 , which is plotted in figure 5 , and for the rebrightening phase it is ccf@xmath61 as plotted in figure 6 . any correlation with ccf above the limit value is likely to be due to the real signal .
we employed the area centroid of the ccf function above the limit value to derive the time lag ( koen .
the uncertainty of time lag was obtained by applying a mc simulation described above , with considering the light curve uncertainties , to compute the distribution of time lags .
all the uncertainties in our results are given but the @xmath62 level of confidence . during the first peak ( figure 5 ) , the soft x - ray and optical emission are positively correlated with maximum ccf of 0.85 , which has a 5.0 @xmath58 level of confidence . the soft x - ray lags optical emission by [email protected] days . during the rebrightening stage ( figure 6 ) ,
the soft and hard x - ray emission are correlated with maximum ccf of 0.61 ( 4.3@xmath58 level of confidence ) , they vary simultaneously with no obvious lag ( time lag = @xmath64 days ) . the soft x - ray and optical emission are weakly correlated with maximum ccf of 0.54 ( 3.5@xmath58 level of confidence ) , soft x - ray lags optical emission by 2.10 @xmath63 0.28 days . the hard x - ray and optical emission are correlated with maximum ccf of 0.61(4.3@xmath58 level of confidence ) , hard x - ray lags optical emission by @xmath65 days .
the luminosities of the quasi - simultaneous optical and x - ray ( 2 - 10 kev ) observations are shown in figure 7 .
the optical - x - ray detection pairs were used when ( 1 ) there is a x - ray detection within one day of an optical detection , and ( 2 ) the confidence level for the x - ray detection is larger than @xmath62 .
we adopted the approximation l@xmath66 to estimate the optical luminosity ( we are approximating the spectral range to the central wavelength of i band ) the same as russell et al.(2006 ) did .
the optical absorbed flux were dereddened using the extinction map from sumi ( 2004 ) .
luminosities were calculated with the distance @xmath67 kpc ( natalucci et al .
2000 ) .
the disk will be heated by the friction between adjacent layers of the disk when the materials were accreted ( shakura & sunyaev 1973 ) , we chose the relation @xmath68 for the intrinsic thermal emission of a viscously heated disk ( frank et al .
2002 , russell et al .
2006 ) . the disk can be also heated by the x - ray and ultraviolet ( uv ) emission from the inner disk . in this case , @xmath69 was taken for optical emission from x - ray reprocessing ( van paradijs & mcclintock 1994 ) , where @xmath70 is the orbital separation .
the flat optically - thick spectrum of the jets of neutron star can also be extented to the optical regime , we adopted the relation @xmath71 ( migliari & fender 2006 ) for this process .
the theoretical relations for the viscously heated disk model and jet model are plotted on figure 7 , and the normalization @xmath72 for each model was taken from russell et al ( 2007 ) .
the orbital separation of sax j 1810.8 - 2609 is not determined at present , we can not compare the data with the x - ray reprocessing model directly .
russell et al ( 2006 ) got the best fit of @xmath73 for neutron star lmxbs in the hard state , most of which are consistent with x - ray reprocessing model .
the russell-2006-best - fit line was plotted on figure 4 as an indicator of contribution from x - ray reprocessing .
furthermore , in figure 7 , we label the data points at different stages using different symbols . during the initial rise stage ,
only the upper limit of x - ray luminosity was available which gave no tight constraint .
the luminosity ratios at the first peak and the rebrightening stage are similar , while that at the rebrightening stage is slightly higher .
we can see that the jet emission model can hardly contribute such bright optical emission , while both the viscously heated disk model and the russell-2006-best - fit will suffice
. considering the cross - correlation result , it is likely that the optical emission is a combination of emissions with different origins .
the viscously - heated - disk emission will cause the x - ray emission delay the optical emission by a viscous timescale .
the x - ray and optical emission will be simultaneous or the optical lags x - ray by a few seconds for the x - ray reprocessing and jet mechanism . while doing cross - correlation between light curve @xmath74 and @xmath75 , if @xmath74 is a combination of two flux - nearly - equal light curves with different lags ( @xmath76 and @xmath77 ) to @xmath75 , it will give out a lag between @xmath76 and @xmath77 ( zhu & zhang .
2010 ) . at the first peak ,
the @xmath0-day x - ray delay is not significant , which suggests that the optical emission may be a combination of emissions from the viscously heated disk and x - ray reprocessing .
each component contributes a large portion of the optical light .
the x - ray delay becomes significant , reaching to @xmath1 days at the rebrightening stage , which suggests that viscously - heated - disk emission becomes dominant .
the prevail of viscously - heated - disk emission should result from two sides : first , as the luminosity ratio of the optical to the x - ray becomes slightly higher at rebrightening stage , there should be an enhancement of radiation efficiency of viscously heated disk .
it was believed the outburst will have a rebrighteing stage if the disk was only partly irradiated , part of the disk was in shadow at the first peak ( truss et al .
the first peak is caused by the accretion of gas within the irradiation portion of the disc , while the subsequent peaks are caused by the accretion of the gas in the outer disk , until the whole disk is kept in high - viscosity state .
it is natural that the radiation area of the disk at rebrightening stage will be larger , and the temperature may be higher .
second , the x - ray reprocessing emission should be suppressed .
it was affected by the structure of both the x - ray source and the outer disk , if the x - ray source is elevated above the disk or the outer disk is warped , it will have high reprocessing efficiency .
when the x - ray source becomes aligned with the disk plane , the reprocessing efficiency will decrease . here
the change of x - ray delay between the first peak and rebrightening stage is significant , the source should evolve outer disk ( maybe also inner disk ) variation during the outburst .
the @xmath1 days x - ray delay during rebrightening stage properly reflects the timescale of the disk during hard state .
it also reflects the outside - in disk accretion process , as we discuss in the outburst morphology section , the variation of x - ray flux follows the optical properly . compared to the @xmath78 days x - ray / optical delay of gx 339 - 4 during rebrightening stage when the source has entered soft state , the @xmath1 days x - ray delay is short .
sax j1810.8 - 2609 should has a very small accretion disk .
in addition , the source stayed in hard state during rebrightening , it should distribute some time to fill the adaf region within the @xmath79 days . an extremely small disk and small companion star
are expected , which is actually what people expected for a faint sxt
as they sorted sax j1810.8 - 2609 ( heise et al . 1999 ; king 2000 ) . in quiescence
, we only have the detect limits for the optical flux available : m@xmath80 mag and m@xmath81 mag ( greiner et al . 1998 ) .
they correspond to m@xmath82 and m@xmath83 with d = 4.9 kpc ( natalucci et al .
2000 ) , respectively .
if the source follows the same luminosity ratio during quiescent state and hard state , we need go roughly 2.2 magnitude deeper in i - band to detect the optical counterpart during quiescence with respect to the viscously heated disk model .
the optical counterpart in quiescent state is extremely faint consistent with the very small disk .
with the ogle and moa light curves , we have identified the optical counterpart to sax j1810.8 - 2609 during outburst .
two outbursts , occurring in 1998 and in 2007 , were covered by ogle , and the 2007 august outburst was monitored by both moa and ogle . except the two outbursts ,
there is a possibly short - time optical flare without x - ray counterpart at mjd = 52583.5 , which may imply a magnetic - loop reconnection in the outer disk as proposed by zurita et al ( 2003 ) the initial x - ray rise of the 2007 outburst detected by _
rxte_/asm is @xmath84-day later than the optical detection by ogle / moa .
however , the flux - detection threshold of _ rxte_/asm played a critically role to this kind of x - ray delay , we have shown that an exponential extrapolation is also reasonable for the data , which indicates no x - ray delay at the initial rise .
we conclude that no obvious x - ray delay was detected at the initial rise with the present detection threshold .
the optical light curve shows multi - peak , as does the x - ray light curve during the 2007 outburst .
the variation of x - ray follows the optical flux properly , which maps the the outside - in disk accretion process . during the first peak phase
, the soft x - ray emission lags the optical emission by @xmath0 days , the x - ray reprocessing should contribute significantly to the optical flux during the first peak .
there is an obvious @xmath1-day x - ray delay during the rebrightening stage which suggests the viscously - heated - disk emission becomes the dominate optical source .
the x - ray reprocessing emission usually dominates the optical flux of the neutron star lmxb at low / hard state ( russell et al .
2006 ) which is probably the case for the first peak of sax j1810.8 - 2609 .
the rebrightening stage is something unusual . unlike aquila x-1
, the outer disk of sax j1810.8 - 2609 should undergo physical environment variations , the physical properties at inner disk may also be different at the first peak and rebrightening stage , even the source stayed in low / hard state during the outburst .
the @xmath1 days time lag implies a very small - radii disk and fairly small companion star .
the non - detection of the source by either ogle or moa at the limit of @xmath85 mag ( @xmath86 mag ) is consistent with this conclusion .
ogle and moa have been observing the galactic bulge for over a decade .
ogle has published a catalogue with @xmath87 variable stars based on ogle ii data ( wozniak et al , 2002 ) .
the ogle iii data will be available soon .
these datasets may contain a large number of potential rebrightening events due to lmxbs , and will provide an unique opportunity for lmxbs study in the future .
we thank the anonymous referee for very helpful and constructive comments to improve the paper significantly .
this work was supported in part by nsf under ast-0908878 .
we are grateful to martin c. smith for promoting this collaboration .
l.z . thanks ian a. bond for sharing the moa data , jeffrey e. mcclintock and lijun gou for helpful discussion and suggestions .
s near mjd = 50800 - 50900 and the dark blue symbol
s near mjd = 54300 - 54400 mark the outbursts at 1998 march and 2007 august , respectively .
the ogle team observed 4 higher - than - average - flux points outside the outburst phase , and among these only the measurement labeled with red circle has a good chance of representing a true higher flux .
( see text . ) , width=528,height=326 ] 54323 - 54337 ) .
the red dotted line at ccf=0.7 shows the @xmath59 confidence level of the ccf based on the observed uncertainties in the data . the red symbol shows the ccf area centroid above @xmath59 level of confidence . ]
54337 - 54400 ) .
the red dotted line at ccf=0.5 shows the @xmath59 level of confidence .
the red dashed and red dot - dashed short line show the ccf area centroid of optical/_rxte _ and optical/ _ swift _ separately .
the yellow solid shows the ccf area centroid of _
swift_/_rxte_. ] barthelmy , s. d. , et al .
ssrv , 2005 , 120 , 143 bond , i. a. , et al .
mnras , 2001 , 327 , 868 buxton , m. m. & bailyn , c. d. , 2004 , 615 , 880 cannizzo , j. k. , 1994 , 435 , 389 fiocchi , m. 2009 , , 693 , 333 frank , j. , king , a. , & raine , d. j. ( ed . ) 2002 , accretion power in astrophysics ( third edition ; cambridge : cambridge univ . press ) greiner , j. , et al .
iauc , 1998 , 6985 , 1 heise , j. , et al .
apl&c , 1999 , 38 , 297 koen , c. mnras , 2003 , 344 , 798 king , a. r. mnras , 2000 , 315 , 33 jain , r. k. , 2001 , 554 , 181 jonker , p. g. , et al .
mnras , 2004 , 349 , 94 lin , d. , et al . , 2007 , 667 , 1073 levine , a. m. , et al . , 1996 , 469 , l33 maitra , d & bailyn , c. d. , 2008 , 688 , 537 van paradijs , j. , & verbunt , f. 1984 , in aip conf .
115 , a comparison of soft x - ray transients and dwarf novae ( melville , ny : aip ) van paradijs , j & mcclintock , j. e. a&a , 1994 , 290 , 133 wozniak , p. r. , et al .
ycatp , 2002 , 005005101 wozniak , p. r. , aca , 2000 , 50 , 421 zurita , c. , casares , j & shahbaz , t. , 2003 , 582 , 369 zhu , l & zhang , s. n , scchg , 2010 , 53s,196 | in this paper we report the long - term optical observation of the faint soft x - ray transient sax j1810.8 - 2609 from ogle and moa .
we have focused on the 2007 outburst , and also did the cross - correlate between its optical light curves and the quasi - simultaneous x - ray observations from _ rxte / swift_. both the optical and x - ray light curves of 2007 outburst show multi - peak features .
quasi - simultaneous optical / x - ray luminosity shows that both the x - ray reprocessing and viscously thermal emission can explain the observed optical flux .
there is a slightly x - ray delay of @xmath0 days during the first peak , while the x - ray emission lags the optical emission by @xmath1 days during the rebrightening stage , which suggests that x - ray reprocessing emission contributes significantly to the optical flux in the first peak , but the viscously - heated - disk origin dominates the optical flux during rebrightening .
it implies variation of the physical environment of the outer disk , even the source stayed in low - hard state during the whole outburst .
the @xmath1-day x - ray lag indicates a small accretion disk of the system , and the optical counterpart was not detected by ogle and moa during quiescence , which constrained it to be fainter than m@xmath2 = 7.5 mag . there is a suspected short - time optical flare detected at mjd = 52583.5 without x - ray counterpart detected
, this single flux increase may imply a magnetic loop reconnection in the outer disk as proposed by zurita et al ( 2003 ) .
the observations cover all stages of the outburst , however , due to the low sensitivity of _
rxte_/asm , we can not conclude whether it is an optical precursor at the initial rise of the outburst . |
the european space agency s gaia mission , approved for launch in 201012 , aims at surveying the galaxy to 20th visual magnitude , using a combination of astrometric measurements ( for trigonometric parallaxes and proper motions ) , multiband photometry ( for basic stellar parameters like temperature and metallicity ) , and radial - velocity measurements .
targeted accuracies versus magnitude allow direct distances and motions to be obtained for large samples of intrinsically bright stars across the galaxy and in some nearby local group galaxies .
expected typical accuracies are shown in table 1 . in total more than 1 billion stars will be observed , of which 50100 million will obtain individual parallax distances to better than 5 per cent .
a primary science goal is to study formation , evolution and structure of the galaxy , for which large - scale mappings of star formation histories are essential . for a full description of the very broad range of science goals
see perryman et al .
( 2001 ) . in its present design gaia
comprises two astrometric instruments , with @xmath4 m@xmath5 apertures and a combined 0.5 deg@xmath5 field of view , and a separate photometric / spectroscopic instrument with a @xmath6 m@xmath5 aperture .
the latter performs photometry in @xmath711 bands for astrophysical classification , and @xmath8 spectroscopy in the 849874 nm wavelength range , mainly for radial velocities . during its lifetime of at least 5 years
, the satellite will scan the entire sky repeatedly , so that each object is observed at multiple epochs .
the above numbers and accuracy predictions refer to the recently ( may 2002 ) completed revised design , aiming at a substantially reduced mission cost compared with the previous baseline ( perryman et al .
2001 ) , while preserving all science goals intact .
cccccccc @xmath9 & @xmath10 & @xmath11 & @xmath12 & @xmath13 & @xmath14 & @xmath15)$ ] & @xmath16 + mag & @xmath17as & @xmath17as yr@xmath18 & km s@xmath18 & & & & kpc + 15 & 13 & 8 & 1.1 & 0.007 & 0.20 & 0.24 & 25 + 17 & 32 & 18 & 6.3 & 0.01 & 0.27 & 0.32 & 60 + 19 & 90 & 50 & & 0.04 & 0.60 & 0.63 & 150 + 20 & 160 & 90 & & 0.13 & 1.1 & 1.3 & 250 +
the availability of precise photometry is essential for age derivations using isochrone fitting to the main sequence turn - off ( msto ) point .
simulations of gaia photometry demonstrate that this method may be successfully exploited with gaia even in such distant stellar systems as the magellanic clouds ( kuinskas et al .
2002 ) , but only for populations younger than @xmath19 gyr . in this paper
we argue that gaia observations of agb stars can be used to determine star formation histories to even greater distances and for much older populations .
gaia will provide a wealth of astrometric and spectrophotometric data on galactic and extragalactic agb stars .
their uses are at least twofold : ( a ) as kinematic tracers , using distances and space motions obtained from the astrometric and radial - velocity data ; ( b ) for age determinations , using basic stellar - atmosphere parameters ( @xmath0 , @xmath1 and @xmath20 $ ] ) derived from the spectrophotometric data , combined with distances and theoretical isochrones . from the astrometric and radial - velocity accuracies in table 1
it is obvious that gaia will yield accurate distances ( @xmath21% ) and full space velocities ( @xmath22 km s@xmath18 ) for individual agb stars up to distances of @xmath23@xmath24 kpc , if no interstellar extinction is present .
extensive simulations by the vilnius gaia group ( vanseviius et al . 2002 ; kuinskas et al .
2002 ) show that gaia will also provide precise metallicities ( @xmath15 ) \leq 0.3 $ ] ) and gravities ( @xmath25 ) for agb stars brighter than @xmath26 ( table 1 ) .
precise effective temperatures ( @xmath27 ) are derived down to @xmath28 .
this holds within a broad range of metallicities ( @xmath20>-2 $ ] ) and ages ( 0.0515 gyr ) .
metallicity estimates of intermediate age and old stellar populations can also be obtained from the slope of the red giant branch ( e.g. ferraro et al .
our simulations show that the method could provide an independent estimate of @xmath29 $ ] with gaia , effective up to distances of @xmath30 kpc , if no interstellar extinction is present ( kuinskas et al .
2002 ) .
we have recently shown ( kuinskas et al .
2000 ) that reliable ages can be derived using isochrone fits to the agb sequences on the observed hr diagram .
it is essential for this procedure to have precise effective temperatures of the agb stars , which can be derived by fitting synthetic spectral energy distributions to observed photometric fluxes ( e.g. , _ bvrijhk _ ) .
the method was successfully tested and compared with the msto method on a sample of populous star clusters in the magellanic clouds spanning a wide range of ages ( table 2 and fig . 1 ) .
for galactic agb stars , it is clear that the distance information needed to construct the observational hr diagrams will be available through gaia .
it thus appears that precise age estimates ( @xmath31 ) can be obtained for a wide range of ages ( 0.0510 gyr ) and metallicities ( @xmath20>-2 $ ] ) .
gaia will provide unique astrometric and photometric data for studying individual and collective properties of stars in the galaxy and its surroundings .
agb stars , being intrinsically bright , will provide precise individual distances , kinematics , @xmath0 , @xmath1 and @xmath20 $ ] up to distances of @xmath2315 kpc .
using isochrone fitting to the agb stars will give reliable ages ( @xmath32 ) for a wide range of ages and metallicities .
if distances are known by other means ( e.g. in distant clusters ) , the method can be used up to @xmath33 kpc .
thus , agb stars will allow the formation histories and kinematics of stellar populations to be probed in a diversity of astrophysical environments both in the milky way and in neighbouring galaxies .
the work was supported by a grant ( nb00-no30 ) of the nordic council of ministers and by a grant of the wenner - gren foundations .
ak thanks the workshop organisers for financial support to attend the event .
bertelli , g. , bressan , a. , chiosi , c. , fagotto , f. , nasi , e. : 1994 , _
a&as _ , * 106 * , 275 .
ferraro , f.r .
, montegriffo , p. , origlia , l. , fusi pecci , f. : 2000 , _ aj _ , * 119 * , 1282 .
katz , d. , munari , u. : 2002 , _ gaia rvs status report , rvs - coco-004 _ + ( ` http://wwwhip.obspm.fr/gaia/rvs/bibliography/rvs-coco-004.txt ` ) .
kuinskas , a. , vanseviius , v. , sauvage , m. , tanab , t. : 2000 , in : mid- and far - infrared astronomy and future space missions , eds .
t , matsumoto & h. shibai , _ isas report sp _
14 * , 51 .
kuinskas , a. , bridius , a. , vanseviius , v. : 2002 , _ ap&ss _ , * 280 * , 159 .
perryman , m. a. c. , de boer , k. s. , gilmore , g. , hog , e. , lattanzi , m. g. , lindegren , l. , luri , x. , mignard , f. , pace , o. , de zeeuw , p. t. , 2001 , _ a&a _ , * 369 * , 339 .
vanseviius , v. , bridius , a. , drazdys , r. : 2002 , _ ap&ss _ , * 280 * , 31 . | we discuss the tracing of star formation histories with esa s space astrometry mission gaia , emphasizing the advantages of agb stars for this purpose .
gaia s microarcsecond - level astrometry , multi - band photometry and spectroscopy will provide individual distances , motions , @xmath0 , @xmath1 and @xmath2 $ ] for vast numbers of agb stars in the galaxy and beyond .
reliable ages of agb stars can be determined to distances of @xmath3200 kpc in a wide range of ages and metallicities , allowing star formation histories to be studied in a diversity of astrophysical environments . -20pt
-20pt -20pt -10pt |
the two sources hercules a ( 3c 348 ) and mrc 0625 - 536 are radio galaxies classified as
fanaroff - riley i ( fr i ) type objects , although hercules a has indeed the morphology of a fr i radio galaxy , but with a much higher radio luminosity , typical of a fr ii object .
the galaxy associated with mrc 0625 - 536 is the main member of the cluster a 3391 , while hercules a lies in a group .
recent spectral and morphological analyses of x - ray data , collected by rosat and asca ( otani et al .
1998 , siebert et al . 1999 ,
gliozzi et al .
2000 ) , have provided information on the physical and dynamical properties of the environment surrounding these objects and of their active nuclei .
both radio galaxies are embedded in diffuse clouds of x - ray emitting hot plasma of size larger than the extended , radio emitting , regions .
the active nucleus may also contribute to the x - ray flux , as observed in several fr i radio galaxies , and this emission is likely related to the inner region of the relativistic jet ( worrall & birkinshaw 1994 , trussoni et al .
1999 , hardcastle & worrall 1999 ) .
in this paper we present bepposax observations of hercules a and mrc 0625 - 536 .
this analysis provides new information on the structure of these radio galaxies and their associated clusters .
in fact , the instrumentation on board of bepposax allows us to collect data , during the same pointing , that range from the soft band up to @xmath5 kev .
we will also examine the variation of the spectral parameters of the x - ray emission across the clusters , relying upon the smoother ( with respect to the asca detectors ) radial behavior of the point spread function of the bepposax mecs detectors . in section 2
we summarize the known properties of the two sources , in section 3 we outline the details of the observations and of the data analysis while in section 4 we report the results of the spectroscopic analysis .
their implications on the structure of the two sources are discussed in detail in section 5 , taking into account also previous observations , and in section 6 we summarize our results .
a value of @xmath6 km s@xmath7 mpc@xmath7 is assumed throughout .
we list in table 1 the main properties of hercules a and mrc 0625 - 536 and recall in the following the available literature . [ cols="^,^,^,^,^,^,^,^,^,^,^,^,^",options="header " , ] @xmath8 @xmath9 ; quantities without errors are fixed @xmath10 photons @xmath11 s@xmath7 kev@xmath7 ( at 1 kev ) the flux detected in the mecs in the whole region ( @xmath12 in radius ) is fit to a thermal emission model with @xmath13 kev and @xmath14 , assuming @xmath15 @xmath11 ( see fig .
the total unabsorbed flux is @xmath16 erg @xmath11 s@xmath7 , with luminosity of @xmath17 erg @xmath11 s@xmath7 , in both the 2 - 10 kev and 0.1 - 2.4 kev bands . after selecting four concentric circular regions with spacing @xmath18 we carried out fits to the emitted fluxes deriving values of the temperature and metallicity basically consistent with the ` average ' values obtained analyzing the whole cluster region ( see table 4 and fig .
we obtain a similar value for the temperature @xmath19 setting @xmath20 in the fit . in the innermost region we can notice an increase in temperature and a decrease in metallicity , that may be a signature of non thermal emission from the active nucleus . to test this point further , we have verified that the central flux also fit a power law model with @xmath21 ( @xmath22 ) .
the total number of photons from this zone ( @xmath23 ) does not allow a complete analysis with a composite spectrum .
setting @xmath24 kev , @xmath25 and @xmath26 we obtain acceptable fits ( @xmath27 ) with fluxes of the two components ( 2 - 10 kev , with large errors ; see table 4 ) @xmath28 erg @xmath11 s@xmath7 and @xmath29 erg @xmath11 s@xmath7 , respectively .
this corresponds to a nuclear luminosity of @xmath30 erg s@xmath7 ( a similar value of @xmath31 is obtained in the softer band 0.1 - 2.4 kev ) .
little useful information can be obtained from the lecs observation : setting @xmath20 we derive @xmath32 kev and @xmath33 @xmath11 ( @xmath34 , 12 d.o.f . ) .
this value is larger than the galactic column density by a factor @xmath35 , however its lower limit ( at 2@xmath36 level ) , @xmath37 @xmath11 , is coincident with @xmath38 .
the flux detected in the high energy band by the pds ( at @xmath39 level ) is well above the extrapolated thermal emission from the cluster and from the active nucleus , assuming the parameters derived from the mecs observation .
in fact , the fluxes expected in the 15 - 200 kev energy range are @xmath40 erg @xmath11 s@xmath7 and @xmath41 erg @xmath11 s@xmath7 , respectively , below the detection capability of the pds ( see fig .
the values of the temperature and metallicity obtained by the bepposax observations are fully consistent with those of asca and pspc ( otani et al .
1998 ) within @xmath42 statistical fluctuations .
the total luminosity from the cluster in the 2 - 10 kev range is @xmath43 lower than derived from asca consistently with the slightly smaller region of photon extraction assumed for the mecs .
the nuclear luminosity derived from the hri observation ( gliozzi et al .
1999 ) exceeds by a factor @xmath44 the one predicted by the correlation of canosa et al .
( 1999 ) , but is still within 90% of uncertainty .
this would suggest the absence of heavy local absorption in the nucleus of this radio galaxy .
however , the mecs data indicate a much higher luminosity ( by a factor @xmath45 6 ) with respect to hri , but , due to the large statistical fluctuations , the lower limit of the luminosity in the mecs is @xmath46 three times the one in the hri . assuming no variability , this discrepancy could be ascribed to a local column density of @xmath47 @xmath11 surrounding the core . this point will be discussed further in the next section , considering also the results of the pds observation .
we outline the main implications of the bepposax observations considering also previous observations from other satellites .
our results confirm the discrepant value of the temperature of the thermal diffuse emission obtained from the asca and rosat observations and support the possibility that a two temperature plasma is present in hercules a. however , since both mecs and pspc data are consistent with emission from an isothermal medium , it is difficult to gain some insights about the origin of these two thermal components .
unfortunately it is not possible to merge the lecs and mecs data to perform a spectral fit with a two temperature model ( see section 3 ) .
some clues for the interpretation of this phenomenology may be provided by the hri image ( siebert et al .
1999 ) , that shows peculiar morphologies , unrelated to the extended thermal gas and suggested to originate from the interaction of the expanding radio lobes with the hot medium .
we can then argue that this processed gas has physical conditions far from those of a plasma in ionization equilibrium . on the other hand ,
the extent of this irregular structure is smaller but comparable with the size of the diffuse region modeled with a @xmath48 profile ( @xmath49 , see fig . 4 of siebert et al .
1999 ) , such that the imaging capability of the mecs does not allow to spatially disentangle these different regions .
it is worth noticing that the luminosity and temperature of the thermal plasma detected in hercules a are consistent with the correlation found for the galaxy clusters ( wu et al .
therefore , the x - ray properties of this object are typical of rich clusters , despite the fact that hercules a has been associated with a group of galaxies , which would be expected to show a lower temperature and luminosity ( zirbel 1997 , ponman et al .
1996 ) .
the mecs data confirm also that only a small fraction of emission from the central region can originate from the core , that however it is very bright when compared with similar objects .
it turns out that in fact that the x - ray luminosity of the nucleus of hercules a is typical of that of fr ii narrow lines radio galaxies ( @xmath50 w hz@xmath7 sr@xmath7 , see fig . 6 in hardcastle & worrall 1999 )
this may mean that extra contributions to the core emission could come from other components , e.g. from an thick accretion disk ( we can not exclude also that the inner nucleus is actually obscured by the torus and the non thermal x - ray emission originate from the jet on much larger scales ) .
on the other hand , the optical and radio core luminosities are consistent with the properties of fr i radio galaxies where obscuration does not play a relevant role and no thick accretion disks appear to be present ( chiaberge et al . 1999 ) .
however the above argument implies that in this radio galaxy the low value of the core radio luminosity , with respect to the total one , may not be a reliable indicator of local absorption in the nucleus . in conclusion , hercules a appears very peculiar at x - ray energies on large and small scales : observations from the new generation telescopes ( chandra ) will provide new insights on this interesting object .
the mecs data confirm the spectral properties obtained by the previous observations with asca and rosat .
the values of the temperature and metallicity are typical for a cluster , and the luminosity is again consistent with the @xmath51-@xmath19 correlation ( wu et al .
the lack of any variation of the temperature across the region of emission confirms that this cluster has very likely undergone a merging process ( see also otani et al .
1998 ) .
the detection of hard x - ray flux in the pds , even though weak , may have important implications on the interpretation of this radio galaxy .
x - ray emission above 20 kev is commonly observed in seyfert galaxies ( see e.g. matt 1998 ) . among clusters ,
hard x - ray emission has been found so far only in coma and a 2256 with good significance , and in a 2199 with marginal significance ( fusco - femiano et al .
1999 , 2000 , kaastra et al . 1999 ) as stated in sect .
4.2 , the high energy emission can not be the extrapolation of the thermal emission from the intracluster plasma , detected from the mecs .
should this emission originate from a hotter thermal component , the fit to the pds data would require a temperature @xmath52 kev . we can reasonably exclude the presence in this cluster of a plasma with these extreme physical conditions . for investigating the origin of the hard emission , we have first looked for possible contamination by a different source in the pds field of view . the nearby cluster
a 3395 falls within the pds field of view , but its thermal emission could not increase the flux by more than a factor @xmath53 ( henriksen & jones 1996 ) .
a bl - lac object , ms 0622.5 - 5256 , is present in the pds field , at the position ra@xmath54
= 06@xmath55 23@xmath56 39.1@xmath57 , dec@xmath54 = -52@xmath58 57@xmath59 49@xmath60 , i.e. at about 49@xmath59 from the field center .
the flux of this source in the 0.3 - 3.5 kev energy band is @xmath61 erg @xmath11 s@xmath7 , estimated assuming a power - law spectrum with a photon index @xmath62=2 ( stocke et al . 1991 ) . extrapolating this flux to the pds energy band , and correcting for the attenuation due to the distance of the bl - lac from the field center
, it turns out that it is about 100 times lower than the flux detected in the pds and therefore can not be responsible for it , unless an unusually strong variability in the flux of the bl - lac object has occurred .
thus , it seems reasonable to consider that the pds flux is of non - thermal origin , and actually originating from the target itself .
we consider in the following three main possibilities : 1 ) it is related to the nucleus of the radio galaxy mrc 0625 - 536 ; 2 ) it is related to the radio lobes of mrc 0625 - 536 ; 3 ) it is associated to the cluster a 3391 .
\1 ) in this case , the hard x - ray emission would originate from the core of radio galaxy mrc 0625 - 536 .
a similar case is represented by the radio galaxy ngc 1275 ( 3c 84 ) , in the perseus cluster , where hard non thermal emission has been detected by oso7 ( rothschild et al . 1981 ) .
bepposax observations have confirmed the existence of x - ray emission at energies exceeding 20 kev , with flux of @xmath63 erg @xmath11 s@xmath7 in the 20 - 100 kev band ( molendi 1998 ) . according to the mecs analysis and to the results of previous observations ,
the emission from the core of mrc 0625 - 536 is not detectable in the pds energy range .
one might , however , speculate on the existence of a component with a much flatter spectral index : for @xmath64 one would derive a ( 15 - 200 kev ) flux of @xmath65 erg @xmath11 s@xmath7 , corresponding to a luminosity of @xmath66 erg s@xmath7 .
fitting the mecs flux from the central region ( radius @xmath67 ) to a composite spectrum with the same parameters as in table 3 for the thermal component , and with @xmath64 for the power law spectrum , one obtains for the latter a flux @xmath68 erg @xmath11 s@xmath7 in the 15 - 200 kev band , @xmath69 times less than detected in the pds .
however , considering the large statistical fluctuations , the difference between the upper limit of the mecs flux and the lower limit of the pds flux ( 90% of confidence ) reduces to a factor @xmath70 .
a larger discrepancy is found with the hri data on the nuclear emission , that is @xmath71 3 times less than the lower limit of the pds flux .
this difference disappears if some amount of obscuration is present with @xmath72 @xmath11 , as deduced in section 4.2 . in this case
the intrinsic core luminosity in the soft band would increase by a factor @xmath53 .
this interpretation has two major difficulties .
first , the nucleus would result , in the x - ray band , overluminous with respect to its radio power ( see sec .
the second problem is the extreme flatness of the spectrum . in blazars , where the hard x - ray emission is related to the inverse compton process in the relativistic jet
, it is generally found that @xmath73 for luminosities @xmath74 erg s@xmath7 ( and steeper spectra in the @xmath75 energy band ; see fossati et al .
it has been alternatively proposed that advection dominated disks ( adaf ) could contribute to the x - ray core emission of low power radio galaxies ( sambruna et al .
1999 ) , with a very flat spectral emission ( see e.g. allen et al .
2000 ) , but with a luminosity much lower ( a factor @xmath76 ) than appropriate for our target .
\2 ) the hard x - ray emission might be produced by inverse compton scattering of relativistic electrons within the radio lobes with the photons of the cosmic microwave background ( cmb ) , as detected in a few radio galaxies ( e.g. , feigelson et al .
1995 , tsakiris et al .
1996 ) . a non - thermal power law model with @xmath62 = 2 , gives for the present source a flux @xmath46 4 @xmath77 10@xmath78 erg @xmath11 s@xmath7 in the 15 - 200 kev band . using the total radio flux at 4.8 ghz of 1.85 jy ( morganti et al . 1993 ) and the radio spectral index @xmath461 ( otani et al .
1998 ) , the detected x - ray flux would imply a magnetic field of about 0.25 @xmath79 g , i.e. more than an order of magnitude lower than the equipartition value derived from the data published by otani et al .
( 1998 ) . with this low magnetic field
, the relativistic electrons emitting in the radio domain up to 8.87 ghz ( frequency to which the radio spectrum extends , otani et al . 1998 ) would have very high energy , largely sufficient to account for the 15 - 200 kev inverse compton emission .
although there is no firm evidence in the literature that radio sources should be in equipartition conditions , there is also no evidence of such large deviations from it .
therefore , this interpretation seems unlikely .
\3 ) non thermal x - ray emission from the cluster a 3391 would be expected if relativistic electrons present in the intergalactic medium undergo inverse compton scattering with the photons of the cmb .
this kind of emission has been detected by bepposax in the two clusters coma ( fusco - femiano et al .
1999 ) and a 2256 ( fusco - femiano et al .
2000 ) , where the presence of ultra - relativistic electrons in the intracluster medium is directly demonstrated by the existence of diffuse radio halos .
conversely , no diffuse radio emission is detected in the cluster a 3391 , so the presence of ultra - relativistic electrons in the intergalactic medium remains questionable . in any event , if the hard x - ray emission is of inverse compton origin , the magnetic field must be very weak in order to avoid producing an observable radio halo .
a similar conclusion is reached for the cluster a 2199 , where a marginal detection of hard x - rays above the expected thermal emission is reported by kaastra et al .
( 1999 ) , and no diffuse radio halo is observed . for the cluster
a 2199 , kempner & sarazin ( 1999 ) suggest the alternative possibility that the x - ray emission above 20 kev could be due to non thermal bremsstrahlung by a population of suprathermal electrons .
the same interpretation might apply to the present case .
we conclude that , even though reasonable , none of the possibilities that we have analyzed for the interpretation of the pds emission is fully convincing .
more data would be needed to clarify this issue .
we summarize the main results obtained from the bepposax observations and their implications on the properties of these two radio galaxies and their environment . _
hercules a. _ no relevant variation of temperature and metallicity occurs between the inner and outer regions of the cluster .
however a plasma with two different temperatures seems to be present , perhaps related to the interaction of the radio components with intracluster gas .
it must be borne in mind anyway that the structure of this object is quite peculiar : at optical wavelengths it is classified as a group but its x - ray properties are typical of clusters .
the non thermal emission from the active nucleus appears unabsorbed , as expected for this class of radio galaxies , but brighter than expected by its core radio emission .
this could be related to the peculiar nature of hercules a that shows properties of both fr i and fr ii radio galaxies . _ mrc 0625 - 36 . _
the x - ray luminosity and the values of @xmath19 and @xmath79 are consistent with those expected for this kind of clusters .
the lack of strong variation of the temperature and metallicity across the emitting region suggests that no cooling flow is present , while it could be related to a recent merging process .
the nuclear non thermal emission detected in the mecs and in previous missions is consistent with that expected from radio data , without excluding however some amount of local obscuration .
concerning finally the high energy flux detected in the pds , it seems unlikely that it is related to the cluster or to the radio lobes . on the other hand , a nuclear origin for this emission would imply a luminosity of some units of @xmath80 erg s@xmath7 in the hard band .
either this emission is completely independent of the flux detected at lower energies , or , to fit the rosat , asca and bepposax ( mecs ) data , it must have an very flat spectrum .
the authors thank m. capalbi , f. fiore and s. molendi for their suggestions in the treatment of the mecs observations of extended sources , and g. brunetti , r. fanti , m. hardcastle , r. morganti , p. padovani and f. vagnetti for discussions and comments on the manuscript .
we warmly acknowledge the late daniele dal fiume , for his invaluable help in the treatment of the pds data .
this work was partially supported by the italian ministry for university and research ( murst ) , under grant cofin98 - 02 - 32 , and by the italian space agency ( asi ) . | we present bepposax observations of the two fr i type radio galaxies hercules a ( 3c 348 ) and mrc 0625 - 536 in the energy range @xmath0 kev .
data analysis shows that the x - ray flux from hercules a is consistent with a diffuse thermal plasma emitting at @xmath1 kev with a possible , but somewhat uncertain , contribution of a softer component at @xmath2 kev
. the non thermal emission from the active nucleus must be significantly smaller than the thermal one , and no indication of relevant core obscuration by a surrounding torus was detected .
the flux from mrc 0625 - 536 originates form an extended region and has been fitted to a thermal law with @xmath3 kev and with a column density consistent with the galactic absorption .
a spatially resolved spectral analysis does not show a relevant variation of the temperature and the metallicity across the diffuse emission zone .
a non thermal spectral component , related to the nuclear activity , may be present in the innermost region with some possible amount of local obscuration , contributing @xmath4 to the total luminosity .
hard x - ray emission from mrc 0625 - 536 has been detected in the pds ( 15 - 200 kev ) that may be related either to its galactic core or to the intracluster region .
1.5 cm |
the jellium model of a metal surface , introduced by bardeen in 1936 , @xcite is the simplest model which reproduces qualitatively , and sometimes quantitatively , the physical properties of real metal surfaces .
@xcite while in his work bardeen applied an approximated hartree - fock ( hf ) theory for the study of the electronic structure , since the seminal work of lang and kohn ( lk ) @xcite the standard theoretical tool applied to the study of the electronic structure of metal surfaces has been density - functional theory ( dft ) .
@xcite as in the original work of lang and kohn , most of the subsequent investigations have applied the local - density approximation ( lda ) of dft , or some of its semi - local variants ( gga , meta - gga , etc . ) .
this approach has been highly successful , and routinely yields good results for global surface properties such as work functions , surface energies , crystal - structure relaxation and reconstruction , etc .
@xcite at a more basic level , however , some problems still remain to be solved , concerning for instance the asymptotic behavior of the exchange - correlation ( _ xc _ ) potential of the widely used kohn - sham ( ks ) approach to dft . in the lda
, this potential decays exponentially when evaluated in the vacuum region , instead of the expected image - like @xmath7 behavior .
@xcite this qualitative failure of the lda _ xc _ potential translates to a similar failure of the position - dependent _ xc _ energy per particle , @xmath10 , which is defined through @xcite @xmath11 = \int n({\bf r } ) \ ; \varepsilon_{xc } ( { \bf r } ) \ ; d{\bf r } , \label{definition}\ ] ] with @xmath12 $ ] being the _ xc_-energy contribution to the universal energy functional of dft , and @xmath13 representing the electron density .
three aspects of eq .
( [ definition ] ) are worth emphasizing : _ ( i ) _ it represents the basic expression for the lda , in which the exact @xmath10 of an arbitrary inhomogeneous electron system is replaced at each point @xmath14 by that of a homogeneous electron gas at the local density @xmath13 , and for a plethora of generalizations of the lda , @xcite _ ( ii ) _ since @xmath12 $ ] can be split as the sum of exchange ( @xmath15 $ ] ) and correlation ( @xmath16 $ ] ) contributions , one can write @xmath17 , and _ ( iii ) _ the position - dependent _ xc _ energy per particle @xmath10 entering eq .
( 1 ) is not unique .
one can always add to @xmath10 an arbitrary function @xmath18 with the condition that weighted by the electron density @xmath13 integrates to zero .
here we have chosen @xmath10 to represent the interaction between a given electron at @xmath14 and its _ xc _ hole .
the goal of this work is to provide exact analytical and numerical calculations of the exact - exchange @xmath19 for jellium slabs and the semi - infinite ( si ) jellium .
in particular , we analyze the asymptotic behavior of the exact @xmath20 in the vacuum region far away from the surface , and we find that there is a qualitative difference between @xmath21 and @xmath22 : both exhibit an image - like behavior of the form @xmath23 , but with a coefficient @xmath24 that while in the case of jellium slabs is universal and equal to 1/2 in the case of a semi - infinite jellium depends on the density of the bulk material and only approaches 1/2 in the low - density limit .
the results reported here should help to settle the still controversial issue of the asymptotic behavior of the position - dependent _ xc _ energy per particle and ks _ xc _ potential at metal surfaces .
@xcite besides , being the results presented here exact at the exchange level , they should also serve as a benchmark against to which dft _ xc _ calculations could be confronted , and hopefully improved , once reduced to their exchange - only version . in this context , very recently luo _ et al . _
@xcite have used a hf scheme to report self - consistent calculations of the surface energy and work function of jellium slabs .
the rest of the paper is organized as follows : in section ii we present the general theoretical background for both jellium slabs and the semi - infinite jellium , and we derive exact analytical expressions for the position - dependent exchange energy per particle in the vacuum region far away from the surface .
numerical calculations that are valid at all positions , from the bulk region to the vacuum , are reported in section iii .
section iv is devoted to the conclusions .
in the case of jellium slabs , with the discrete character of the positive ions inside the slab being replaced by a uniform distribution of positive charge ( the jellium background ) , the positive jellium density is @xmath25 which describes a slab of width @xmath26 , number density @xmath27 , @xcite and jellium edges at @xmath28 and @xmath29 .
@xmath30 represents the heaviside step function : @xmath31 if @xmath32 and @xmath33 if @xmath34 .
the size of the slab is infinite in the @xmath35 plane .
the jellium slab is taken to be invariant under translations in the @xmath35 plane , so the ks eigenfunctions can be rigorously factorized as follows @xmath36 where @xmath37 and @xmath38 are the in - plane coordinate and wave - vector , respectively , and @xmath39 represents a normalization area in the @xmath35 plane .
@xmath40 are normalized spin - degenerate eigenfunctions for electrons in slab discrete levels ( sdl ) @xmath41 @xmath42 with energies @xmath43 they are the solutions of the effective one - dimensional ks equation @xmath44 \xi _ { i}(z)=0 , \label{ksequations}\ ] ] with @xmath45 being the electron mass .
it is important to remark here that the factorization of the 3d wave - function as proposed in eq .
( [ ksfunctions ] ) is only valid for the case of a _ local _ potential , as is the case of the ks implementation of dft . on the other hand , in the hf approximation
the non - locality of the fock potential introduces a coupling between @xmath38 and @xmath41 quantum numbers @xcite . as a consequence ,
hf numerical calculations are more time - consuming than the ones presented here , either lda , kli @xcite , or oep .
@xcite the local ks potential @xmath46 entering eq .
( [ ksequations ] ) is the sum of two distinct contributions : @xmath47 where @xmath48 is the classical ( electrostatic ) hartree potential , given by @xcite @xmath49 . \label{hartree}\ ] ] here , @xmath50 is the electron number density @xcite @xmath51 where @xmath52 , and @xmath53 is the fermi energy or chemical potential , which in turn is determined from the neutrality condition for the whole system by the condition @xmath54 .
@xmath55 is the nonclassical _ xc _ potential , which is obtained as the functional derivative of the _ xc_-energy functional @xmath56 $ ] : @xcite @xmath57}{\delta n(z ) } \ ; . \label{xcpotential}\ ] ] both for the slab and semi - infinite [ @xmath58 limit of eq .
( [ jellium ] ) ) ] geometries , and as a consequence of the translational symmetry in the @xmath35 plane , eq .
( [ definition ] ) simplifies to @xmath11 = a \int\limits_{-\infty } ^{\infty } dz\,n(z)\,\varepsilon_{xc}(z ) , \label{xcenergy}\ ] ] where @xmath59 is the position - dependent _ xc _ energy per particle at plane @xmath1 . in the case of jellium slabs
, the exchange - only contribution to @xmath12 $ ] ( which is originated in the pauli exchange hole with all other correlation effects excluded ) is known to be given by the following expression : @xcite @xmath60=-2e^2 a \sum_{i , j}^{occ .
} k_{f}^{i } k_{f}^{j } \int\limits_{-\infty } ^{\infty } dz\int\limits_{-\infty } ^{\infty } dz^{\prime}\varphi_{i}(z , z^{\prime } ) \varphi _ { j } ( z^{\prime } , z)f_{ij}(z , z^{\prime } ) , \label{xenergy}\ ] ] where @xmath61 and @xmath62 with @xmath63 being the cylindrical bessel function of first order .
@xcite comparison of eq .
( [ xcenergy ] ) and ( [ xenergy ] ) yields the following expression for the exchange - only contribution to @xmath64 : @xmath65 which can be interpreted as the energy due to the interaction of an electron at @xmath1 and its exchange - only pauli hole . in order to demonstrate that the exact - exchange energy per particle @xmath66 of eq .
( [ 5 ] ) represents indeed the interaction between an electron at @xmath1 and its exact - exchange hole , we appeal to the following expression for the exchange - hole for our slab geometry@xcite @xmath67 which represents the density of the exchange hole at point @xmath68 ( observational point ) due to the presence of an electron located at * r*. owing to the translational symmetry in the @xmath35 plane , without loss of generality we have choose @xmath69 , @xmath68 = ( @xmath70 , @xmath71 ) . using eq .
( [ xhole ] ) , and defining @xmath72 , eq .
( [ 5 ] ) may be rewritten as @xmath73 which justifies the physical interpretation of @xmath4 as the interaction energy of an electron located at @xmath1 and the `` charge distribution '' given by @xmath74 .
equations ( [ xhole ] ) and ( [ physical ] ) can also be used as a sort of alternative definition of the @xmath4 investigated in this work , as they solve the non - uniqueness of @xmath4 which results from its definition through the exchange - only version of eq .
( [ xcenergy]).@xcite in order to obtain the asymptotic behavior of @xmath75 at @xmath76 , we first restrict our analysis to the case where there is one single occupied sdl .
@xcite in this special case , in which @xmath77 , eq .
( [ 5 ] ) yields [ see also eqs .
( [ density ] ) and ( [ 2 ] ) ] : @xmath78 ^{2}}{\sqrt{\rho ^{2}+(z - z^{\prime } ) ^{2 } } } , \label{7}\ ] ] or , equivalently ( see appendix ) : @xmath79 , \label{8}\ ] ] with @xmath80 and @xmath81 being the modified bessel and struve functions , respectively .
@xcite we note that eq .
( [ 8 ] ) is valid for _ all _ @xmath82 both inside and outside the jellium slab . also , the cancellation of @xmath50 which occurs in passing from eq .
( [ 5 ] ) to eq . ( [ 7 ] ) allows for the numerical calculation of @xmath83 for arbitrarily large values of @xmath1 .
[ [ asymptotic - behavior . ] ] asymptotic behavior .
+ + + + + + + + + + + + + + + + + + + + for the slab geometry , it is permissible ( and rigorous ) to take the asymptotic limit @xmath84 although the integral over @xmath85 runs from @xmath86 to @xmath87 this is due to the fact that for a given @xmath1 the main contribution to the integral in eq .
( [ 8 ] ) comes from values of @xmath85 inside the slab ( @xmath88 ) , as @xmath89 decays exponentially one or two @xmath90 s from each jellium edge . at this point
, we define @xmath91 and use the asymptotic expansions of @xmath80 and @xmath81 in the limit @xmath92 we obtain @xcite @xmath93 hence , as @xmath94 , eq .
( [ 8 ] ) yields the following asymptotic behavior : @xmath95 where @xmath96 $ ] and @xmath97 $ ] , mean values being defined here as @xmath98 the @xmath99 dependence of @xmath100 and @xmath101 comes from the self - consistent ks wave - functions @xmath102 , which for a given @xmath26 are different for different values of the slab density dictated by @xmath99 . for details on the derivation of eq .
( [ sdpasymp ] ) from eq .
( [ 8 ] ) , we refer to the appendix . for the general situation where more than one sdl is occupied ,
we obtain the asymptotic limit of eq .
( [ 5 ] ) by using the fact that for @xmath103 _ ( i ) _ the electron density is dominated by the slowest decaying ks orbital , which corresponds to the _ highest occupied _ sdl ( @xmath104 ) , and _
( ii ) _ the numerator of eq . ( [ 5 ] ) is dominated by the term @xmath105 , since all @xmath102 with @xmath106 decay exponentially two or three @xmath90 s from each jellium edge .
hence , in the vacuum region far away from the surface we find : @xmath107 and @xmath108 or , equivalently [ see eq .
( [ 2 ] ) ] : @xmath109 ^{2}}{\sqrt{\rho ^{2}+(z - z^{\prime } ) ^{2}}}\;. \label{15}\ ] ] finally , following the same procedure as in the case of a single occupied sdl , we find : @xmath110 where @xmath111 $ ] and @xmath112 $ ] .
@xcite this result represents a straightforward generalization of the result presented above [ eq .
( [ sdpasymp ] ) ] for the case of one single occupied sdl . at this point , it is interesting to note that the leading contribution to @xmath113 can be easily obtained directly from eqs .
( [ 7 ] ) or ( [ 15 ] ) , by approximating the argument inside the square root by @xmath1 ( in the large @xmath1 limit ) , and using the normalization of the ks orbitals @xmath114 and the identity @xmath115 . by considering a slab of thickness sufficiently large to make the energy spectrum _
, solamatin and sahni @xcite reached the conclusion that far away from the slab the so - called slater potential @xmath116 [ which is twice the exchange energy per particle : @xmath117 decays as @xmath118 , in contrast with the asymptotic structure dictated by eq .
( [ sdpasymp2 ] ) .
this result is , however , not correct due to the fact that for a finite jellium slab ( no matter how thick it is ) the slab _ intrinsic _ discrete spectrum [ corresponding to the eigenvalues @xmath119 entering eq .
( [ ksequations ] ) ] can never be replaced by a continuous one .
equation ( [ sdpasymp2 ] ) leads us to the conclusion that in the vacuum region of a finite jellium slab and at distances from the surface that are large compared to @xmath120 ( which is typically larger than the slab thickness @xmath26 ) , @xmath2 , which is exactly half the corresponding ks exact - exchange potential @xmath121.@xcite hence , as in the case of finite systems,@xcite the slater potential @xmath116 of jellium slabs [ or , equivalently , twice the exchange - energy per particle @xmath0 ] embodies the asymptotics of the ks exchange potential @xmath123 .
in contrast , solamatin and sahni @xcite concluded that in the case of a semi - infinite jellium only _ half _ the slater potential embodies the asymptotics of the ks exchange potential , i.e. , @xmath124 ; but nastos @xcite claimed that @xmath125 , so there is still something remaining to be clarified on this issue .
work along these lines is now in progress.@xcite the exchange energy of a strict two - dimensional ( 2d ) electron gas can be obtained from that of a jellium slab with a single occupied sdl [ eq . ( [ xenergy ] ) with @xmath77 ] , by first performing the one - dimensional non - uniform scaling @xcite @xmath127 and then taking the limit as @xmath128 .
the scaling above preserves the total number of electrons .
noting that for a single occupied sdl the jellium - slab exchange energy takes the following form @xmath60=- \frac{2 \pi e^2 a}{\left ( k_{f}^{1 } \right)^2 } \int\limits_{0}^{\infty } \frac{d \rho}{\rho } \left [ j_{1}(k_{f}^{1}\rho)\right]^2 \int\limits_{-\infty } ^{\infty } dz \int\limits_{-\infty } ^{\infty } dz^{\prime } \frac { n^{\text{slab}}(z ) \ ; n^{\text{slab}}(z^{\prime } ) } { \sqrt{\rho^2+(z - z^{\prime})^2 } } , \label{explicit}\ ] ] where @xmath129 we find : @xmath130 = -n\,\frac { 4}{3 \pi } e^2 \ ; k_f^1 , \label{2dlimit}\ ] ] where @xmath131 represents the total number of electrons .
previously , this scaling limit had been formulated in a different way , resulting in the much generous constraint that the exchange energy per particle in the 2d ( @xmath132 ) limit should be greater than @xmath133 .
@xcite it is interesting to note that the exchange energy functional as given by eq .
( [ explicit ] ) is an _
explicit _ functional of the density , which is only possible in this single occupied sdl case , due to the simple ( invertible ) relation between density and wave - function , as given by eq .
( [ 2ddensity ] ) .
in the general , many sdl occupied case , eq .
( [ 2ddensity ] ) is replaced by eq .
( [ density ] ) , the direct inversion from wave - functions to density is not feasible anymore , and the exchange energy functional is an explicit functional of the ks orbitals , but an _ implicit _ functional of the density , as in eq .
( [ xenergy ] ) .
we note at this point that the exchange energy of a strict 2d electron gas can also be obtained directly from eq .
( [ 7 ] ) through the replacement @xmath134 with @xmath135 being the dirac delta function , and taking @xmath29 : @xmath136^{2 } = -n\,\frac{4}{3\pi } e^2 k_{f}^{1}. \label{11}\ ] ] either from eq . ( [ 2dlimit ] ) or ( [ 11 ] )
, we find for the exchange energy per particle of the strict 2d homogeneous electron gas the well - known result @xmath137 .
@xcite in the case of a semi - infinite jellium , a half - space filled with a uniform distribution of positive charge ( the jellium background ) , the jellium density is @xmath138 with the jellium edge at @xmath29 defining the surface of a metal .
as in the case of jellium slabs , the semi - infinite jellium is invariant under translations in the @xmath35 plane , so the ks eigenfunctions can be factorized as follows @xmath139 where @xmath37 and @xmath38 are the in - plane coordinate and wave - vector , respectively , and @xmath39 ( @xmath140 ) represents a normalization area ( length ) .
@xmath141 are spin - degenerate eigenfunctions for electrons with a continuous energy spectrum @xmath142 ( @xmath143 is a continuum quantum number ) .
they are the solutions of the effective one - dimensional ks equation @xmath144 \xi _ { k_z}(z)=0 .
\label{ksequations.semi}\ ] ] the ks potential @xmath46 is given by eq .
( [ kspotential ] ) , as in the case of a jellium slab but with the slab electron density of eq .
( [ density ] ) being replaced by the si electron density @xmath145 in the case of a semi - infinite jellium , the position - dependent exchange energy per particle at plane @xmath1 is given by the following expression : @xmath146 where @xmath147 , and @xmath148 is of the form of eq .
( [ 2 ] ) but with @xmath149 ( @xmath150 ) being replaced by @xmath151^{1/2}$ ] ( @xmath152^{1/2}$ ] ) . [
[ asymptotic - behavior.-1 ] ] asymptotic behavior .
+ + + + + + + + + + + + + + + + + + + + the derivation of the asymptotic limit of @xmath153 is more delicate than in the case of jellium slabs , due to the fact that in the present case we have a continuous energy spectra .
hence , the crucial argument that we have used to derive the asymptotic behavior for jellium slabs , concerning the fact that in that case the highest occupied sdl dominates in the vacuum region far away from the surface , is not so transparent when the spectrum is continuous .
stated in other words , contributions to the asymptotic position - dependent exchange energy of eq .
( [ 5.semi ] ) come indeed from values of @xmath143 and @xmath154 that approach @xmath155 , but not necessarily _ only _ from the highest occupied value , i.e. , from @xmath156 .
the asymptotic behavior of eq . ( [ 5.semi ] ) was analyzed by nastos @xcite by assuming that at @xmath157 the ks potential @xmath46 takes the image - like form @xmath158 , with @xmath159 positive , but otherwise arbitrary .
one finds that in the vacuum region far away from the surface ( @xmath160 ) the ks orbitals @xmath161 can be expanded with respect to the ks orbital at @xmath162 as follows@xcite @xmath163 with @xmath164 @xmath165 standing for the square root of the ratio between the fermi energy and the work function @xmath166 ( @xmath167 ) . by introducing eqs .
( [ aprox.nastos ] ) and ( [ aprox.nastos1 ] ) into eqs .
( [ densidad.semi ] ) and ( [ 5.semi ] ) , one finds@xcite @xmath168 and @xmath169 furthermore , the asymptote of eq .
( [ semi ] ) does not depend on the actual form of @xmath170 .
this is due to a cancellation , when @xmath1 is large , of the orbitals @xmath171 entering the numerator and denominator of eq .
( [ 5.semi ] ) .
this is the reason why the leading term in the expansion of @xmath172 is independent of @xmath159 .
that means that the result remains valid also in the absence of this image - like contribution , i.e. , even assuming that the ks potential @xmath46 entering eq .
( [ ksequations ] ) decays exponentially as @xmath103 .
the asymptotic behavior dictated by eq . ( [ semi ] ) was obtained independently by solamatin and sahni using a somehow less general , but otherwise quite different approach .
@xcite they approximated the ks potential @xmath46 entering eq .
( [ ksequations.semi ] ) by a finite - linear - potential model at the interface region , and used the corresponding orbitals in each of the three regions where the model was defined : trigonometric functions ( in the bulk region ) , airy functions ( near the surface ) , and exponential decaying functions ( in the vacuum ) . as the only thing that matters in obtaining the asymptote of eq .
( [ semi ] ) is the correct expansion of the ks orbitals with respect to the ks orbital at @xmath162 [ as given by eq .
( [ aprox.nastos ] ) ] , and with this general expansion being fulfilled also within the finite - linear - model potential used in refs . and , solamatin and sahni obtained eq .
( [ semi ] ) which is valid in general .
it is also worth of address the fact that the result of eq .
( [ semi ] ) is in contrast with the asymptotic behavior @xmath173 that one obtains for the exchange energy per particle in the case of the airy edge electron gas .
@xcite this is due to the fact that the asymptotic behavior of the solutions of the airy edge gas is different from the one given by eq .
( [ aprox.nastos ] ) , as for this model the potential increases linearly with distance in the vacuum region , instead of approaching a constant value .
an analysis of the so - called pauli and lowest - order correlation - kinetic components of the exchange energy per particle @xmath174 can be found in ref . .
all the numerical calculations presented below have been carried out by ignoring all correlation effects ( beyond the pauli exchange ) . by this
we mean that the _ xc _ potential @xmath55 entering eq .
( [ ksequations ] ) has been replaced by the exchange - only contribution @xmath175 , disregarding @xmath176 , both for jellium slabs and for the semi - infinite jellium . in the case of jellium slabs , @xmath123 and the corresponding ks orbitals of eq .
( [ ksequations ] ) can be obtained through the solution of the discrete version of the @xmath177-only optimized effective potential ( oep ) method , as given for example by eqs .
( 14 ) or ( 20 ) of ref . .
this code is feasible and we have at our disposal these self - consistent exact - exchange ( oep ) ks orbitals and exchange potentials .
@xcite alternatively , an approximate way to obtain @xmath123 and the corresponding ks orbitals ( the exchange - only lda orbitals ) is to replace the actual exchange potential @xmath175 by the exchange potential of a uniform electron gas at the local electron density @xmath178 , i.e. , @xmath179^{1/3}$ ] [ hartrees ] . fig .
1 shows a comparison of ( i ) the well - known lda exchange energy per particle @xmath180^{1/3}$ ] [ hartrees ] , with @xmath178 being the self - consistent electron density obtained with the use of either exchange - only lda orbitals ( wide solid line ) or exact - exchange ( oep ) orbitals ( wide dotted line ) , with ( ii ) the exact - exchange energy per particle @xmath181 of eq .
( [ 5 ] ) obtained by using , as before , either exchange - only lda orbitals ( solid line ) or exact - exchange ( oep ) orbitals ( dotted line ) .
it is important to note that while both alternative evaluations of @xmath182 fail badly in the vacuum region where the actual exchange energy per particle exhibits an image - like asymptotic behavior , the use of lda orbitals in eq .
( [ 5 ] ) results in an exchange energy per particle ( solid line ) that on the scale of the figure is nearly identical to the exact fully - self - consistent result ( dotted line ) .
small differences introduced by the use of lda orbitals ( see the inset ) are mostly localized in the bulk region near the surface , where friedel - like oscillations appear to be too weak in this approximation .
the numerical methods that we have used to obtain exact - exchange ( oep ) orbitals and exchange - only lda orbitals suffer from instabilities in the vacuum region far from the surface . as in this region exchange
- only lda orbitals are stabler than their oep counterparts and the results presented in this work do not depend significantly on whether exact - exchange ( oep ) or exchange - only lda orbitals are used in eq .
( [ 5 ] ) ( see fig . 1 ) , all the calculations presented below have been obtained with the use of exchange - only lda orbitals .
we emphasize , however , that this is not a crucial approximation , and that the magnitude of the error that it introduces is given by the almost indistinguishable difference between the full and dotted lines in fig . 1 . the numerical self - consistent calculations presented in fig . 1 and in the remaining of this section ( which are all obtained from either eq . ( [ 5 ] ) or eq . ( [ 5.semi ] ) with the use of exchange - only lda orbitals )
have been performed as follows . for jellium slabs ,
two infinite barriers have been located in the vacuum region far enough away from the two surfaces , in such a way that all the numerical results be independent of their precise location , @xcite and the ks equations have been solved through a straightforward discretization in real space along the one - dimensional coordinate @xmath1 . in the case of the semi - infinite jellium ,
the ks equations have been solved by following the general procedure introduced by lang and kohn .
@xcite this consists of defining three regions for the solution of the ks equation : far - left ( bulk region ) , central ( a few @xmath183 s to the left and to the right of the jellium edge ) , and far - right ( vacuum region ) . in the bulk region ,
the ks eigenfunctions are taken to be of the form @xmath184 , where @xmath185 are phase shifts , and this fixes an overall normalization constant . in the central region , we define a mesh of @xmath186 points between @xmath187 and @xmath188 @xmath189 , the first point @xmath187 being chosen far enough from the jellium edge in the bulk so that the friedel oscillations can be neglected , and the outer point @xmath188 being chosen to be far enough from the jellium edge into the vacuum so that the effective one - electron potential is negligibly small . since @xmath190 for @xmath191 , the orbitals can be approximated as @xmath192 where @xmath24 is a constant and @xmath193 .
the ks orbitals at the mesh points are calculated by using the numerov integration procedure .
@xcite as in the vacuum region the orbitals follow exponential form , it is numerically most stable to integrate them inwards , so the numerov integration procedure in this case is given by : @xmath194 where @xmath195 .
finally , matching the ks orbitals in the central region with the corresponding analytical expression in the bulk region [ @xmath196 determines the constant @xmath24 . in fig . 2
, we consider a thin jellium slab with @xmath197 ( corresponding to the average electron density of al ) and @xmath198 .
this slab contains _
one _ single occupied sdl , so that we compare our full numerical calculation of eq .
( [ 5 ] ) ( solid line ) with the asymptote of eq . ( [ sdpasymp ] ) ( dashed - dotted line ) .
we see that @xmath181 reaches eq .
( [ sdpasymp ] ) at about one @xmath183 from the jellium edge and reaches the asymptote @xmath199 at a few fermi wavelengths ( @xmath200 ) from the surface . in fig . 3
, we consider a jellium slab with @xmath197 and @xmath201 .
for this particular case , _ nine _ sdl s are occupied , i.e. , @xmath202 , so we compare our full numerical calculation of eq .
( [ 5 ] ) ( solid line ) with the asymptote of eq . ( [ sdpasymp2 ] ) ( dashed - dotted line ) .
the main message of this figure is that _
( i ) _ in the vacuum region far away from the surface @xmath181 is dominated by the term @xmath105 ( dashed - dotted - dotted line ) , and _
( ii ) _ @xmath181 reaches the asymptote @xmath199 only at a distance from the jellium edge of several fermi wavelengths ( @xmath203 ) .
we remind here that point _
( i ) _ above was our main assumption in the derivation of the slab asymptotic limit of eq .
( [ sdpasymp2 ] ) .
this assumption is fully justified after the numerical results shown in fig .
3 . at this point , with a few algebraic manipulations , we rewrite the asymptote of eq .
( [ sdpasymp2 ] ) in the physically motivated image - like form @xmath204 } , \label{image.like}\ ] ] where @xmath205 and @xmath206 , which represents the location of the so - called image plane , results in @xmath207 ^ 2}{z } + o \left ( \frac{1}{z^{2 } } \right ) .
\label{zeta.cero}\ ] ] as @xmath103 , @xmath206 reaches a finite value , given by @xmath208 in fig . 4
, we plot a comparison of eq .
( [ zeta ] ) ( solid line ) with the image - plane position that we obtain by fitting our full numerical calculation of eq .
( [ 5 ] ) with the image - like eq . ( [ image.like ] ) and @xmath209 ( empty circles ) .
for this , we have used a fit region of width @xmath210 centered at @xmath211 from the jellium edge in the vacuum .
differences between eq .
( [ zeta ] ) ( solid line ) and our numerical estimate ( empty circles ) , which in the case of very thin films are negligible , are entirely due to the fact that the fitting of the numerical calculation must be carried out in a vacuum region that extends very far away from the surface .
empty squares correspond to the result of eq .
( [ zeta.cero ] ) , including the correction to the leading term , and taking @xmath212 ( which is the average value of the fit region indicated above ) .
according to eq .
( [ zeta ] ) , @xmath213 is inversely proportional to @xmath214 , which exhibits an oscillatory behavior as a function of the slab width @xmath26 ( see the solid line in the upper part of fig . 4 ) going to zero every time a sdl becomes occupied .
hence , the location of the image plane becomes infinitely negative every time a sdl becomes occupied , which results in the strong finite - size oscillations shown in fig .
now we focus on a comparison between our full numerical jellium - slab and semi - infinite - jellium calculations of the position - dependent exchange energy per particle ( see fig . 5 ) . in the bulk , as @xmath26 increases the slab calculations converge with the semi - infinite calculation ( see the inset at the upper part of fig . 5 ) , and both slab and semi - infinite calculations approach in the bulk region far away from the surface the exchange energy per particle of a three - dimensional ( 3d ) homogeneous electron gas , @xmath215 . in the vacuum , however , there is always a region far enough away from the surface where the jellium slab and the semi - infinite jellium behave differently : while all slab calculations converge to an image - like behavior of the form of eq .
( [ image.like ] ) with @xmath209 , the semi - infinite @xmath216 exhibits an image - like behavior in agreement with eq . ( [ semi ] ) ( see fig .
5 also shows that as the width @xmath26 increases the slab @xmath75 coincides with the semi - infinite @xmath217 in a wider vacuum region near the surface ( see the lower inset of fig .
6 shows a comparison of our full numerical calculation of eq .
( [ 5.semi ] ) with the asymptote of eq .
( [ semi ] ) ( solid and dashed lines , respectively ) for @xmath197 . in this case , @xmath218 ( as obtained from our exchange - only lda self - consistent calculation of the work function @xmath166 ) and eq .
( [ semi ] ) yields @xmath219 ( dashed line ) , which is in contrast with the asymptote of eq .
( [ sdpasymp2 ] ) ( dotted line ) that holds in the case of jellium slabs . the same comparison has been done for other values of @xmath220 , and we have found that our full numerical calculation is always very close ( as in fig .
6 ) to the asymptote of eq .
( [ semi ] ) . finally , we display in fig . 7 our exchange - only lda self - consistent calculation of the coefficient @xmath221 in a wide range of electron densities .
it is important to note that the corresponding coefficient @xmath222/\left[2 \pi ( 1+\alpha^2)\right]$ ] ( see the inset to fig .
7 ) entering eq .
( [ semi ] ) is close to @xmath223 at metallic densities ( @xmath224 ) .
7 also shows that only at extremely low densities the coefficient @xmath165 approaches zero , thereby the coefficient @xmath222/\left[2 \pi ( 1+\alpha^2)\right]$ ] of eq .
( [ semi ] ) approaching @xmath8 .
hence , the asymptotic limits of @xmath75 and @xmath217 only coincide in the low - density limit ( @xmath225 ) .
we have presented a detailed analysis of the position - dependent exchange energy per particle @xmath0 at jellium slabs and the semi - infinite jellium . for jellium slabs , we have found that in the vacuum region far away from the surface @xmath226 , independent of the bulk electron density .
this is the equivalent to the well - known result @xmath227 , which holds in the case of localized finite systems like atoms and molecules .
@xcite the equivalence between these results is however not straightforward , since slabs have an extended character in the @xmath35 plane , being `` localized '' only along the @xmath1-coordinate . in the vacuum side of the surface
there is a region where @xmath181 coincides with @xmath153 and this region increases as @xmath26 increases .
the fitting of our numerical calculations of @xmath181 to a physically motivated image - like expression is feasible , but the resulting location of the image plane @xmath228 $ ] shows strong finite - size oscillations . in particular , we have shown analytically that @xmath229 , @xmath230 being a signature of the energy of the highest occupied sdl with respect to the fermi level . for a semi - infinite jellium
, we have found that our numerical calculations agree well with the analytical asymptote [ see eq .
( [ semi ] ) ] obtained in refs . and , which approaches the slab asymptote @xmath231 only in the extreme low - density limit ( @xmath225 ) .
we attribute the qualitatively different behavior of @xmath232 and @xmath172 to the fact that these asymptotes are approached in different ranges .
while in the case of the semi - infinite jellium the asymptote is reached at distances @xmath1 from the surface that are large compared to the fermi wavelength ( the only existing length scale in this model ) , for slabs the asymptote is reached at distances @xmath1 from the surface that are large compared to @xmath120 ( which is typically larger than the slab thickness @xmath26 ) . for thick slabs with @xmath233 ( @xmath183 being the fermi wavelength ) , @xmath181 first coincides with @xmath153 [ dictated by eq .
( 36 ) at @xmath234 in the vacuum region near the surface ( see fig . 5 ) , but at distances from the surface that are large compared to @xmath120 , @xmath181 turns to the slab image - like behavior of the form of eq . ( 22 ) [ or , equivalently , eq . ( 38 ) with @xmath235 ; in the limit as @xmath236 ( i.e. , when the jellium slab becomes semi - infinite ) , @xmath237 _ coincides _ with @xmath238 _ everywhere_. in the low - density limit , where @xmath239 , the condition @xmath233 is never fulfilled and @xmath181 reaches ( at @xmath240 ) one _ single _ asymptote : the slab image - like behavior of the form of eq .
( 22 ) [ or , equivalently , eq . ( 38 ) with @xmath235 , which turns out to coincide with the semi - infinite - jellium asymptotic behavior dictated by eq .
( 36 ) .
finally , we note that as @xmath241 , the same conclusion is expected to be valid for the exchange contribution to the position - dependent _ xc _ energy per particle .
recent developments concerning the asymptotic behavior of the correlation contribution to the ks exchange correlation potential @xmath55 of a semi - infinite jellium can be found in ref . .
c.m.h wishes to acknowledge the financial support received from conicet of argentina .
c.r.p . was supported by the european community through a marie curie iif ( mif1-ct-2006 - 040222 ) and conicet of argentina through pip 5254 .
j.m.p . acknowledges partial support by the university of the basque country , the basque unibertsitate eta ikerketa saila , and the spanish ministerio de educacin y ciencia ( grants no . fis2006 - 01343 and csd2006 - 53 ) .
here we will shown in detail how to go through eqs .
( [ 7])-([sdpasymp ] ) in the text . starting from eq .
( [ 7 ] ) , one first notice that@xcite @xmath242 \ ; , \ ] ] with @xmath243 and @xmath244 being the modified bessel and struve functions , respectively .
substitution of eq .
( [ a1 ] ) in eq .
( [ 7 ] ) yields at once eq .
( [ 8 ] ) .
now , in the asymptotic limit,@xcite @xmath245 which inserted in the definition of the function @xmath246 yields eq .
( [ 9 ] ) .
substitution of this expansion for @xmath246 in eq .
( [ 8 ] ) leads to the expression , @xmath247 \ ; , \\ = -\frac{e^2}{2 } \int\limits_{-\infty}^{\infty}dz^{\prime}\frac{|\xi_1(z^{\prime})|^2}{|z - z^{\prime}| } + \frac{e^2}{\pi k_f^1}\int\limits_{-\infty}^{\infty}dz^{\prime}\frac{|\xi_1(z^{\prime})|^2}{|z - z^{\prime}|^2 } -\frac{e^2}{4(k_f^1)^3}\int\limits_{-\infty}^{\infty}dz^{\prime}\frac{|\xi_1(z^{\prime})|^2}{|z - z^{\prime}|^4 } + ... \;. \label{a4}\end{aligned}\ ] ] expanding the denominators of eq .
( [ a4 ] ) in the large @xmath1 limit , the different contributions in eq .
( [ sdpasymp ] ) arise . for instance , the leading contribution @xmath248 comes from the first term on the r.h.s . in eq .
( [ a4 ] ) , with @xmath249 approximated by @xmath1 .
it is interesting to note that one important feature of this result , as is its material and slab - size independence , is consequence ( in this context ) of the normalization of the sdl wave - functions . on more general grounds , and returning to the alternative definition of @xmath181 given in eq .
( [ physical ] ) , this is more physically understood as a consequence of that the integral of @xmath250 over all possible `` observational '' coordinates ( @xmath251 ) is exactly @xmath252.@xcite the next term in the expansion , proportional to @xmath253 and with decay @xmath254 , is obtained from the sub - leading contribution of the first term on the r.h.s . in eq .
( [ a4 ] ) , together with the leading contribution from the second term . to the order explicitly displayed in eq .
( [ sdpasymp ] ) , no contribution arises from the last ( third ) term in eq .
( [ a4 ] ) , as the leading contribution coming from this term to @xmath255 is of the order @xmath256 .
while this analysis has been performed for the single - occupied sdl case , it also applies to the general case where more than a sdl is occupied , as explained in section ii.a.2 .
@xmath257 is given by @xmath258 , with @xmath220 being the dimensionless electron - density parameter defined as the radius of a sphere containing on average one electron and @xmath259 being the bohr radius .
a convenient length unit for the present system is the fermi wavelength @xmath260 .
t. grabo , j. kreibich , s. kurth , and e. k. u. gross , in _ strong coulomb interactions in electronic structure calculations : beyond the local density approximation _ , edited by v. i. anisimov ( gordon and breach , amsterdam , 2000 ) .
this hartree potential includes the contribution coming from the uniform positive background ( proportional to @xmath261 ) .
alternatively , this contribution could have been denoted separately as the `` external potential '' . due to the imposed translational invariance in the @xmath35 plane ,
functional derivatives here are conveniently defined as @xmath263/\delta g(z)]\delta g(z ) dz$ ] , where @xmath264 represents a uniform variation of the function @xmath265 in the plane @xmath1 .
this is the origin of the factor @xmath266 in eq .
( [ xcpotential ] ) .
most of the calculations presented here have been found to be well converged by locating the two infinite barriers at @xmath276 from each jellium edge . for the case of one single occupied sdl , the two infinite barriers have been taken at @xmath277 from each jellium edge . | the position - dependent exact - exchange energy per particle @xmath0 ( defined as the interaction between a given electron at @xmath1 and its exact - exchange hole ) at metal surfaces is investigated , by using either jellium slabs or the semi - infinite ( si ) jellium model . for jellium slabs ,
we prove analytically and numerically that in the vacuum region far away from the surface @xmath2 , _ independent _ of the bulk electron density , which is exactly half the corresponding exact - exchange potential @xmath3 [ phys .
rev .
lett . * 97 * , 026802 ( 2006 ) ] of density - functional theory , as occurs in the case of finite systems .
the fitting of @xmath4 to a physically motivated image - like expression is feasible , but the resulting location of the image plane shows strong finite - size oscillations every time a slab discrete energy level becomes occupied . for a semi - infinite jellium , the asymptotic behavior of @xmath5 is somehow different . as in the case of jellium slabs
@xmath6 has an image - like behavior of the form @xmath7 , but now with a density - dependent coefficient that in general differs from the slab universal coefficient @xmath8 . our numerical estimates for this coefficient agree with two previous analytical estimates for the same .
for an arbitrary finite thickness of a jellium slab , we find that the asymptotic limits of @xmath4 and @xmath5 only coincide in the low - density limit ( @xmath9 ) , where the density - dependent coefficient of the semi - infinite jellium approaches the slab _ universal _ coefficient 1/2 . |
for many years x - ray astronomers have promised to obtain accurate constraints on dark matter in clusters of galaxies and elliptical galaxies .
but because of the frustrating limitations of previous x - ray telescopes , only for a very few objects notably m87 have precise measurements been possible .
it is really a great pleasure to give this review because the promises made many years ago are finally being realized in this wonderful era of x - ray astronomy , where the _ chandra _ and _ xmm _ observatories are operating so successfully .
_ chandra _ and _ xmm _ have provided for the first time high quality , spatially resolved spectra of the diffuse hot gas of galaxies and clusters because their ccds combine moderate resolution spectra with very much improved spatial resolution and sensitivity .
_ chandra _ provides a more significant jump in spatial resolution while xmm provides a more substantial boost in sensitivity . as a result of these improved capabilities , accurate measurements of the gas temperature as
a function of radius exist for many clusters .
these measurements provide very interesting constraints on the dm . because most of the published results on x - ray studies of dark matter ( dm ) using _ chandra _ and _ xmm _ exist for clusters , in this review
i will emphasize the results obtained on the radial dm distributions in clusters .
my discussion will be divided up into segments that address the mass distributions inside and outside of cluster cores .
i devote the remaining space to elliptical galaxies , particularly ngc 720 , where i will discuss x - ray constraints on the ellipticity of dm halos .
in galaxy clusters the dominant baryonic component is that of the hot gas , yet it contributes only 10 - 30 percent to the total mass .
clusters are therefore `` dm - dominated '' and are excellent sites to study the properties of the dm .
in fact , in the context of the cdm model , simulations by suggest that clusters are dm - dominated down to less than 1% of the virial radius ( @xmath0 ) , making them especially attractive for studies of the cores of dm halos .
another advantage of studying clusters is that there are a variety of methods that can be used to probe their mass distributions
stellar / galaxy dynamics , gravitational lensing , and dynamics of the hot ( x - ray ) gas .
each of these methods has certain advantages and disadvantages .
the x - ray method , which is the focus here , is primarily hampered by the assumption of hydrostatic equilibrium and questions about the thermodynamic state of the hot gas .
as for the assumption of hydrostatic equilibrium , provided one selects clusters with regular morphologies , hydrodynamic simulations show that the x - ray method is robust , even if the cluster is not in perfect equilibrium @xcite .
further support for hydrostatic equilibrium is indicated by the generally good agreement between cluster masses obtained from weak lensing and x - rays @xcite , though some disagreements with strong lensing remain @xcite . regarding the state of the hot gas , mass determinations using x - ray data usually assume the gas is single - phase .
indeed , the _ chandra _ and _ xmm _ observations of clusters have justified that assumption ( e.g. , ; ) .
these observations have shown that outside cluster cores the hot gas is single - phase . however , within cluster cores the data are consistent with , but do not necessarily require , a limited multiphase medium with a temperature range much less than that expected from a standard multiphase cooling flow . in a few of the cooler systems
there is clear evidence for multiphase gas in their cores ( m87 , centaurus , ngc 5044 ) .
although the single - phase assumption certainly appears valid outside cluster cores , the situation is less clear within the cores of cool clusters and groups . for a single - phase gas in hydrostatic equilibrium the calculation of the radial mass distribution from x - ray data
is fairly straightforward .
typically , one assumes spherical symmetry and divides up the x - ray image into a series of concentric , circular annuli .
then coronal plasma models are fitted to the annular spectra to infer the temperature and density of the gas as a function of radius .
often this procedure is done by first deprojecting the data using an `` onion peeling '' method pioneered by andy fabian and collaborators .
then parameterized functions are fitted to the radial profiles of the gas density and temperature to allow analytical calculation of the derivatives in the hydrostatic equation .
the effects of rotation and magnetic fields are usually ignored but should be negligible @xcite .
data from next - generation x - ray satellites should provide the first interesting constraints on gas rotation .
let us consider first the results obtained with _
chandra _ and _ xmm _ for the mass profiles of clusters outside their cores .
perhaps the most impressive example is that of a1413 studied by .
this bright cluster has a very regular x - ray image , the isophotes are moderately flattened , and the radial surface brightness profile shows no structure other than a central enhancement over a single @xmath1 model .
these authors obtained an excellent determination of the temperature profile between 100 kpc and 1.5 mpc ( see figure [ fig.a1413 ] ) .
pratt & arnaud determined the mass profile by examining different parameterizations for the temperature and density of the hot gas .
outside their central data point , they found the gravitating mass profile is very precisely constrained and is quite insensitive for the specific parameterizations of the gas density and temperature .
they find that the nfw and moore et al.profiles provide good fits to the shape of the gravitating mass profile from @xmath2 out to @xmath3 and give reasonable values for the concentration parameter , @xmath4 ( for nfw ) .
a1835 is another example of a generally relaxed cluster .
obtained good constraints on the mass profile using _ chandra _ data .
their approach to constrain the mass is a slight variation on the procedure mentioned above .
they start with an assumed mass profile and use the well - constrained gas density profile to predict the radial gas temperature profile which is then fitted to the data .
schmidt et al .
find acceptable fits to their temperature profile out to 1 mpc @xmath5 using an nfw model and obtain a reasonable value for the concentration @xmath6 .
they also find that their mass profile agrees with that obtained from a strong lensing analysis .
_ chandra _ results for a2390 ( figure [ fig.a2390 ] ) another lensing cluster have been presented by .
this cluster is not as regular as those just discussed .
instead , there is some substructure in the core and irregularities in the x - ray image on @xmath7 kpc scales .
the radial surface brightness profile even shows a curious break near a radius of @xmath8 mpc .
hence , this is certainly a cluster where the assumption of hydrostatic equilibrium is suspect .
however , allen et al .
find that the nfw model is a good fit to the data and gives a reasonable value for the concentration @xmath9 .
very good agreement is also observed between the x - ray determined mass and that obtained from strong and weak lensing studies .
thus , whatever departures from hydrostatic equilibrium exist in this cluster do not appear to translate to large errors in the derived mass profile .
in addition to the `` hot '' clusters ( @xmath10 kev ) just described , interesting constraints on the mass profiles have also been constrained for cooler clusters .
the _ xmm _
data of the cool @xmath11 kev ) cluster a1983 have been analyzed by .
the x - ray image and radial surface brightness profiles are quite regular .
the temperature profile of the hot gas is well constrained ( and slightly rising ) out to @xmath8 mpc .
pratt & arnaud do find that the nfw profile is a good fit to the data , but the inferred concentration parameter @xmath12 is lower than expected ( @xmath13 ) for this low - mass cluster . they argue that departures from hydrostatic equilibrium are likely the cause of the discrepancy , however , the regularity of the x - ray image does not provide obvious support for that possibility .
the authors also find marginal evidence for multi - temperature gas in the central region which should also be considered a possible origin for the anomalous concentration value .
many more clusters also have been analyzed with _
chandra _ and _ xmm _ observations ; e.g. , a2199 @xcite , 3c295 @xcite , and ms1008.1 - 1224 @xcite .
overall , nfw ( and moore et al . )
profiles are found to be good fits to the mass profiles outside cluster cores ( @xmath14 @xmath15 ) obtained from _
chandra _ and _ xmm _ observations .
exceptions ( particularly a1983 ) are probably the result of departures from hydrostatic equilibrium and/or multi - temperature gas .
cdm predicts that dm halos should have approximately a power - law form at small radius @xmath16 , though the precise value of the power - law exponent remains controversial . for cdm , simulations indicate that @xmath17 is between 1 ( nfw ) and 1.5 ( moore et al . ) . but
observations of lsb galaxies indicate @xmath18 in disagreement with cdm .
this discrepancy is the motivation of the `` self - interacting '' dm model ( sidm ) proposed by .
it is natural to ask whether clusters also have @xmath18 since they are dm - dominated deep into their cores .
of course , to address this issue we need to focus on clusters with relaxed cores where the assumption of hydrostatic equilibrium will be valid . unfortunately , _
observations have shown that nature has provided a very serious roadblock to using x - ray observations to study the mass profiles in cluster cores .
clusters that tend to be the most relaxed systems on hundred - kpc and mpc scales have usually been associated with `` cooling flows '' . but
observations with _
chandra _ have shown that in the central regions of cooling flows ( i.e. , within @xmath19 kpc ) the x - ray surface brightness is highly disturbed .
the hot gas exhibits filaments and holes where these holes are often filled by radio emission from an agn in the central galaxy .
such morphological irregularities clearly call into question the assumption of hydrostatic equilibrium .
attempts to measure the core mass profile using _ chandra _ observations of clusters with such obvious disturbances have found disagreement with cdm .
although obtain @xmath20 in the hydra - a cluster , they obtain a concentration parameter @xmath21 that is higher than expected . in the core of a1795 find @xmath22 in clear disagreement with cdm .
it is difficult to evaluate these results for hydra - a and a1795 since departures from hydrostatic equilibrium could easily explain the discrepancies with cdm .
clearly , it is essential to find x - ray clusters that are relaxed in their cores .
some such clusters have now been identified by considering those systems that are known to be be relaxed on hundred - kpc and mpc scales but also do not have a strong radio source in the central galaxy .
one such cluster is a2029 ( figure [ fig.a2029 ] ) which _ chandra _ shows is indeed regularly shaped all the way down into its core @xcite .
this massive cluster is also nearby ( @xmath23 ) and very bright , making it an ideal target for x - ray analysis of its core mass distribution .
spectra of a2029 extracted within thin annuli from the center out to @xmath24 kpc are well - fitted by a single temperature component .
this is a key result since one may have expected possible complications from a multiphase gas in the center of a cooling flow .
these authors find that the mass profile of a2029 is reasonably well described by a simple power law that is quite insensitive to different parameterizations of the measured gas density and temperature profiles .
the exponent of the power - law fit to the gravitating mass profile is @xmath25 .
this translates to @xmath26 for the mass density profile , very similar to the nfw / moore values , yet very significantly larger than the values @xmath18 obtained for lsb galaxies .
the nfw model is a better fit to the mass profile than either a power - law or the moore et al .
model though a fit of similar quality is obtained for the hernquist profile .
the inferred concentration parameter is reasonable @xmath27 . after subtracting out the contribution to the gravitating mass from the hot gas and stars
, it is found that the dm has very similar properties to that of the gravitating matter : @xmath28 for a power - law mass model and @xmath29 for nfw .
( note : these results quoted for a2029 include an extra data point at large radius not used by and thus differ slightly from the previously published results . ) in we present a _
observation of a2589 which is another cluster that is found to be regularly shaped all the way into into its core .
our _ chandra _ observation unfortunately only yielded 8 ks of good data , and thus the constraints are not as precise as those obtained for a2029 . like a2029 ,
the hot gas in a2589 is consistent with a single - phase gas .
the gas density and temperature are well constrained and described by simple functions the gas is consistent with being isothermal .
the results for the gravitating mass profile are quite consistent with those obtained for a2029 .
a power - law fit to the mass profile yields @xmath30 for the mass density .
a better fit is provided by nfw which is a good fit down to @xmath31 with a reasonable concentration parameter @xmath32 .
similar results are obtained for the dm : @xmath33 for a power - law mass model and @xmath34 for nfw . similar agreement with cdm
is also observed for some clusters @xcite that are not quite so obviously relaxed in their cores as a2029 and a2589 but are clearly not as disturbed as systems like hydra - a and a1795 .
hence , in contrast to the clusters with unrelaxed cores , the core dm profiles inferred for clusters with relaxed cores appear to be quite consistent with cdm .
unfortunately , since the x - ray halos of individual elliptical galaxies are much fainter than those of clusters and groups of galaxies , it is much more difficult to obtain interesting constraints on the gravitating mass in elliptical galaxies using x - rays .
in fact , only for ngc 720 has a detailed analysis of the mass distribution using _ chandra _ data been published so far .
the ellipticities of dm halos provide important constraints on the nature of the dm in cosmological models essentially independent of @xmath35 or @xmath36 @xcite .
nearly spherical halos are expected for the cores of dm halos if the dark matter is self - interacting @xcite .
highly flattened ( nearly disk - like ) dm halos occur in models where the dm is in the form of cold , molecular gas @xcite .
halos formed in a @xmath36cdm universe are moderately flattened with typical ellipticities @xmath37 , though they are rounder if there is significant `` warm '' dm @xcite .
in contrast to the case of spherical symmetry discussed so far , there is no simple analytic solution of the equation of hydrostatic equilibrium for ellipsoidal mass distributions .
however , the added complexity of ellipsoidal symmetry does offer some subtle rewards .
for a single - phase gas in hydrostatic equilibrium the volume x - ray emissivity and the gravitational potential have identical shapes independent of the temperature profile ( an `` x - ray shape theorem '' ) .
this allows a robust `` geometric test '' of the hypothesis that mass follows the optical light .
one just deprojects the stellar image and computes the potential assuming mass follows the optical light .
the shape of the potential is then compared to the shape of the deprojected x - ray image independent of the gas temperature profile . this test for dm has been applied to a _
observation of the nearby ( @xmath38 mpc ) galaxy ngc 720 @xcite .
this is an isolated e4 galaxy that is fairly bright in x - rays .
chandra _ image of this galaxy revealed over 40 discrete sources embedded within the diffuse emission within @xmath39 of the galaxy center .
these sources , most of which were undetected by previous observations , are wonderful for studying the properties of x - ray binaries , but are a nuisance for studies of dm .
these sources were removed from the image and replaced with estimates of the local diffuse emission .
the image of ngc 720 cleaned of discrete sources displays x - ray isophotes with moderate flattening and orientations that are slightly offset from the optical .
starting at the center of the galaxy , the positional angles ( pas ) of the isophotes slowly twist @xmath40 away from the optical major axis out to @xmath41 kpc .
there is a further twist at larger radii .
here we focus on the region within @xmath42 kpc . if the gas is in hydrostatic equilibrium , since this twist does not occur in the stellar image it implies the existence of a massive dm halo according to the geometric test mentioned above .
the x - ray ellipticity profile is consistent with a constant value ( @xmath43 ) out to @xmath42 kpc at which point it increases .
again , we restrict our discussion to the region within this radius .
the x - ray ellipticities are much less than the stellar values ( @xmath44 ) .
but they are larger than predicted by the geometric test ( figure [ fig.n720 ] ) .
the mass - follows - optical - light hypothesis is rejected using ellipticities alone at slightly more than the @xmath45 level .
the evidence for dm provided by the x - ray pa twist and ellipticities can not be explained away by mond .
first the need for dm occurs already in the region where newtonian gravity should dominate .
second , the potential shapes in mond are the same as in newtonian theory . note that gas rotation , which is not accounted for in this analysis , is not expected to be important for ngc 720 @xcite .
we find that the ellipticity of the gravitating mass distribution , which is dominated by the dm , is not very sensitive to the mass model ( figure [ fig.n720 ] ) .
we find @xmath46 for @xmath47 and @xmath48 for nfw .
note that if the stars make a sizable contribution to the potential ( which is not supported by the _ chandra _ data ) then the ellipticity of the dm will be larger than that of the gravitating matter .
the dm ellipticity inferred for ngc 720 is consistent with cdm predictions but is very inconsistent with the round halos predicted by sidm and the very flat halos predicted if the dm is cold gas .
the pa twist , particularly within @xmath42 kpc strongly suggests the dm is triaxial the twist appears to agree with the triaxial model of ngc 720 proposed by . although these results give nice agreement with cdm predictions , it is essential to verify them using independent methods and with other galaxies .
the stellar velocity field could offer such an independent test since the strongly triaxial mass distribution indicated by the x - ray data should produce a highly asymmetric stellar velocity field ( t. statler , private communication ) .
we are currently analyzing new observations of the elliptical galaxies ngc 1332 and ngc 3923 which will provide key comparisons to mass properties of ngc 720 .
we emphasize that ngc 720 shows the importance of detecting and accounting for discrete sources embedded in the diffuse emission .
argue that the _ chandra _ measurements of highly flattened x - ray isophotes of the elliptical galaxy ngc 1700 indicate that rotation must be dynamically important in the hot gas .
however , these authors detect only 2 - 3 discrete sources in their system , and the x - ray ellipticity profile they derive matches the optical ellipticities over almost the whole radius range investigated .
this strongly suggests that that the morphology of the x - ray emission of ngc 1700 is strongly biased by the unresolved discrete sources .
presently , the key result obtained from _
chandra _ and _ xmm _ studies of dm in clusters and ellipticals is that the radial density profiles in galaxy clusters and the ellipticity of the dm halo ( for ngc 720 ) agree well with cdm predictions . in particular , the dm profiles measured deep down into the cores of the relaxed galaxy clusters a2029 and a2589 rule out an important contribution from self - interacting dm of the kind proposed to account for dm in the cores of lsb galaxies .
i thank the conference organizers for the invitation to provide this review and for arranging such an excellent meeting .
i also thank my collaborators on x - ray studies of dm for their contributions to some of the papers discussed in this review . | i review constraints on the radial density profiles and ellipticities of the dark matter obtained from recent x - ray observations with _
chandra _ and _ xmm _ of elliptical galaxies and galaxy clusters and discuss their implications , especially for the self - interacting dark matter model .
# 1_#1 _ # 1_#1 _ = # 1 1.25 in .125 in .25 in |
it remains an open question whether massive neutrinos are majorana particles , whose antiparticles are themselves @xcite . the final answer to this fundamental question will tell us whether the lepton number is conserved or not in nature , and help us explore the origin of neutrino masses .
currently , the most promising way to answer if massive neutrinos are their own antiparticles is to observe the @xmath0 decays @xmath19 , where @xmath20 and @xmath21 stand respectively for the atomic and mass numbers of a nuclear isotope @xmath22 @xcite . over the last few decades
, a great number of dedicated experiments have been carried out to search for this kind of decays @xcite .
so far , we have not observed any positive signals , and a lower bound on the half - life of the implemented nuclear isotope can be drawn from experimental data .
the gerda phase - i experiment @xcite has disproved the signals of @xmath23 decays claimed by the heidelberg - moscow experiment @xcite , and the joint lower bound from all the previous @xmath24-based experiments on the half - life turns out to be @xmath25 at the @xmath26 confidence level @xcite . for @xmath27-based experiments ,
a combined analysis of the exo-200 @xcite and kamland - zen phase - i data @xcite gives rise to a lower bound @xmath28 at the @xmath26 confidence level .
more recently , kamland - zen announced their phase - ii result @xcite , and improved the lower bound to @xmath29 at the @xmath26 confidence level with both phase - i and phase - ii data .
if neutrino mass ordering is inverted ( i.e. , @xmath30 ) , the next - generation @xmath23 experiments with a few tons of target mass will be able to discover a remarkable signal in the near future @xcite .
the schechter - valle theorem @xcite states that a clear signal of @xmath23 decays will unambiguously indicate a finite majorana mass of neutrinos , if neither a fine - tuning among parameters nor a cancellation among different contributions is assumed .
decay rate remains nonzero as the nuclear medium effects on quarks may break any intricate cancellation .
] obviously , this theorem signifies the physical importance of searching for @xmath23 decays experimentally .
the quantitative impact of the schechter - valle theorem has already been studied by duerr , lindner and merle in ref .
@xcite , where it is found that the majorana neutrino masses implied by the schechter - valle theorem are too small to explain neutrino oscillations .
explicitly , assuming one short - range operator to be responsible for @xmath23 decays , they find that current experimental lower bounds on the half - lives of @xmath23-decaying isotopes indicate an upper bound on the majorana neutrino mass @xmath31 , where @xmath32 denotes the effective neutrino mass term associated with @xmath33 for @xmath34 . in this paper
, we reexamine this problem , and obtain an upper bound @xmath35 that agrees with the above result from ref .
@xcite on the order of magnitude .
furthermore , we generalize the analysis of @xmath0 decays to that of the lnv rare decays of @xmath15 , @xmath13 and @xmath36 mesons .
for instance , we obtain @xmath37 , @xmath38 and @xmath39 from current upper bounds on the lnv rare decays of @xmath36 mesons .
the radiative majorana neutrino masses related to other lnv decays are also tabulated .
therefore , we confirm the conclusion from ref .
@xcite that although the schechter - valle theorem in general implies a tiny majorana neutrino mass , we have to explore other mechanisms to generate the observed neutrino masses at the sub - ev level .
the remaining part of this work is organized as follows . in sec . 2
, we recall the calculation of majorana neutrino masses from the four - loop diagram mediated by the effective operator , which is also responsible for the @xmath0 decays .
the generalization to the lnv meson decays is performed in sec . 3 , where the corresponding majorana masses are computed .
finally , we summarize our main conclusions in sec . 4 .
in this section , we present a brief review on the calculation of majorana neutrino masses radiatively generated from the operator that leads to the @xmath0 decays , following ref .
@xcite closely .
such a calculation can be readily generalized to the case of majorana neutrino masses induced by the lnv meson decays , as shown in the next section . at the elementary - particle level
, the @xmath0 decays can be expressed as @xmath40 , where the up quark @xmath41 , the down quark @xmath42 and the electron @xmath43 are all massive fermions .
if the @xmath0 decays take place , they can be effectively described by the lnv operator @xmath44 , in which the chiralities of charged fermions have been omitted and will be specified later . as already pointed out by schechter and valle @xcite
, this operator will unambiguously result in a majorana neutrino mass term @xmath45 .
the relevant feynman diagrams are given in fig .
[ fig:0n2b ] .
it is worthwhile to notice that quark and charged - lepton masses are indispensable for the schechter - valle theorem to be valid , as emphasized in ref .
@xcite . in the standard model ( sm ) ,
only left - handed neutrino fields participate in the weak interactions , so the electron masses can be implemented to convert the right - handed electron fields into the left - handed ones , which are then coupled to left - handed neutrino fields via the charged weak gauge boson @xmath46 .
this does make sense , since the chirality of electrons in the operator @xmath47 can in general be either left - handed or right - handed . for the same reason , quark masses are also required to realize the hadronic charged - current interactions in the sm . in this case , the operator @xmath48 in fig .
[ fig:0n2b](a ) can be attached to the left - handed neutrinos through two propagators of @xmath46 , leading to the neutrino self - energy diagram in fig .
[ fig:0n2b](b ) . assuming that @xmath0 decays are mediated by short - range interactions
, one can write down the most general lorentz - invariant lagrangian that contains various point - like operators as follows @xcite @xmath50 where @xmath51 and @xmath52 denote respectively the fermi constant and the proton mass , and @xmath53 ( for @xmath54 ) are effective coupling constants . in eq .
( [ eq : operators ] ) , the hadronic currents are defined as @xcite @xmath55 ( 1 \pm \gamma_5^ { } ) d \ ; , \end{aligned}\ ] ] while the leptonic currents are given by @xmath56 ( 1 \pm \gamma_5^ { } ) e^{\rm c } \
; , \end{aligned}\ ] ] where @xmath57 with @xmath58 is the charge - conjugated electron field .
according to @xmath59 and the fact that fermion fields are grassmann numbers , one can immediately verify that the tensor leptonic current @xmath60 automatically vanishes .
different chiralities of hadronic and leptonic currents in eqs .
( [ eq : hadronic ] ) and ( [ eq : leptonic ] ) should be distinguished by the left- and right - handed projection operators @xmath61 .
for instance , we have defined @xmath62 , and similarly for the other types of currents in eqs .
( [ eq : hadronic ] ) and ( [ eq : leptonic ] ) , in which the corresponding subscripts
l " or r " are omitted without causing any confusions . in this connection
, the effective coupling constants @xmath53 should also be regarded as @xmath63 ( for x , y , z = l , r ) , which are carrying the superscripts for different chiralities of hadronic and leptonic currents .
given one of the five operators in eq .
( [ eq : operators ] ) , one can set an upper limit on their coupling @xmath64 by assuming that it is responsible for the @xmath65 decay and saturates the experimental lower bound on the half - life , as done in ref .
recently , some of those limits have been recalculated in ref .
@xcite , using more recent results for the nuclear matrix elements .
the effective coupling constants for the operators @xmath66 and @xmath67 have been found to be @xmath68 and @xmath69 , respectively .
having obtained these couplings , we are then ready to evaluate the induced neutrino mass by inserting the dimension - nine effective operators into the butterfly diagram , as depicted in fig .
[ fig:0n2b ] .
the authors of ref .
@xcite have demonstrated that the operator @xmath66 leads to a vanishing neutrino mass term via the butterfly diagram , while the other one @xmath70 does lead to a tiny majorana neutrino mass , which will be revisited below .
now that the operator @xmath71 is responsible for the @xmath0 decays , the radiatively induced majorana mass term for electron neutrinos can be extracted from the self - energy in fig . [
fig:0n2b](b ) by setting the external momentum to zero as @xcite , @xmath72 where @xmath73 is the weak gauge coupling , and @xmath74 and @xmath75 are the up - quark , down - quark and electron masses , respectively .
in addition , the loop integral is given by @xmath76 ^ 2_{}$ ] with @xmath77 ( q_1 ^ 2 - m_d^2 ) ( k_1 ^ 2-m_{w}^2 ) } \ ; , \end{aligned}\ ] ] where @xmath78 is the @xmath79-boson mass , @xmath80 and @xmath81 stand for the four - momenta of internal particles running in the loop and can be easily identified via the integrand on the right - hand side of eq .
( 5 ) and from fig .
[ fig:0n2b](b ) . to evaluate this integral
, we employ the technique for massive two - loop diagrams in ref .
@xcite and arrive at @xmath82 with @xmath83 as usually introduced in the dimensional regularization and @xmath6 being the renormalization scale .
the relevant function reads as @xcite @xmath84 [ ( p+q)^2 + m_k^2 ] } \\ & = & \pi^4 ( \pi m_i^2)^{n-4 } \frac{\gamma(2-\frac{1}{2}n)}{\gamma(3-\frac{1}{2}n ) } \int_0 ^ 1 { \rm d}x \int_0 ^ 1 { \rm d}y ~ [ x ( 1-x)]^{\frac{1}{2}n-2 } y ( 1-y)^{2-\frac{1}{2}n } \nonumber \\ & & \times \left\ { \frac{(y^2 \kappa^2 + \eta^2)\gamma(5-n)}{[y(1-y)\kappa^2 + y + ( 1-y)\eta^2 ] ^{5-n } } + \frac{n}{2 } \frac{\gamma(4-n)}{[y(1-y)\kappa^2 + y + ( 1-y)\eta^2 ] ^{4-n } } \right \},\end{aligned}\ ] ] with @xmath85 as usual , the integral is expanded with respect to @xmath86 in the limit of @xmath87 and the ultraviolet divergences can be separated as inverse powers of @xmath88 . since the loop integral involves the divergent terms proportional to @xmath89 and @xmath90 , we need to keep the terms up to @xmath91 in @xmath92 , namely , @xmath93 so as to obtain all the finite parts of @xmath94 . in our numerical calculations , we have adopted the renormalization scale of @xmath95 , which is a characteristic scale of typical energy transfer in nuclear processes .
the other implemented parameters can be found in table [ tb : constants ] . in the scheme of minimal subtraction , we finally get the induced neutrino mass from eq .
( [ eq : sigmap ] ) as latexmath:[\[\begin{aligned } which agrees with the result @xmath31 from ref .
@xcite on the order of magnitude .
since this mass is extremely small , one has to implement other mechanisms to account for neutrino masses . in this sense , the main conclusion in ref .
@xcite is still valid that the schechter - valle theorem is qualitatively correct , but quantitatively irrelevant for the neutrino mass - squared differences required for neutrino oscillation experiments . | the schechter - valle theorem states that a positive observation of neutrinoless double - beta ( @xmath0 ) decays implies a finite majorana mass term for neutrinos when any unlikely fine - tuning or cancellation is absent . in this note , we reexamine the quantitative impact of the schechter - valle theorem , and find that current experimental lower limits on the half - lives of @xmath0-decaying nuclei have placed a restrictive upper bound on the majorana neutrino mass @xmath1 radiatively generated at the four - loop level . furthermore , we generalize this quantitative analysis of @xmath0 decays to that of the lepton - number - violating ( lnv ) meson decays @xmath2 ( for @xmath3 , @xmath4 = @xmath5 or @xmath6 ) .
given the present upper limits on these rare lnv decays , we have derived the loop - induced majorana neutrino masses @xmath7 , @xmath8 and @xmath9 from @xmath10 , @xmath11 and @xmath12 , respectively .
a partial list of radiative neutrino masses from the lnv decays of @xmath13 , @xmath14 and @xmath15 mesons is also given .
* majorana neutrino masses from neutrinoless double - beta decays and lepton - number - violating meson decays * * jun - hao liu @xmath16 * + @xmath17institute of high energy physics , chinese academy of sciences , beijing 100049 , china + @xmath18center for high energy physics , peking university , beijing 100871 , china pacs number(s ) : 11.30.fs , 12.15.lk , 14.60.lm |
the phenomenology related to the qcd confinement / deconfinement phase transition is nowadays a fascinating subject in the center of intense investigations , both experimentally and theoretically ( see _ e.g. _
@xcite for a review of the topic ) . during the last two decades , a new state of matter
, the quark - gluon plasma ( qgp ) , has been intensively studied through heavy - ion collisions ( sps , rhic or lhc ) and is still carefully analysed .
the experiments seem to conclude that the qgp behaves like a perfect fluid with a low ratio viscosity over entropy around the critical temperature of deconfinement @xmath1 .
therefore , this observation suggests that a strongly - coupled phase ( called sqgp ) is present in this temperature range and that the qcd confinement / deconfinement phase transition is much more driven by screening effects of the strong interaction . in order to correctly describe the different mechanisms at work during this phase transition , or more exactly during this crossover ,
a lot of theoretical researches ( lattice qcd , phenomenological approaches ... ) are carried out . in particular , finding the qgp equations of state ( eos ) is a crucial information nowadays needed .
the aim of this work is to investigate the thermodynamic features of the qgp by resorting to a phenomenological approach based on @xmath0-matrix computations .
this approach has the advantage to allow the study of bound and scattering states of the system in a whole picture .
relevant results have already been established for heavy - quarkonia above @xmath1 @xcite and also for glueballs in the yang - mills plasma @xcite . moreover , in this latter reference , the eos of the yang - mills plasma for su(n ) and g@xmath4 have been computed thanks to the dashen , ma and bernstein s formulation of statistical mechanics in terms of the @xmath5-matrix ( or @xmath6-matrix ) @xcite .
such a formulation is particularly well suited for systems whose microscopic constituents behave according to relativistic quantum mechanics .
the qgp is indeed identified to a quantum gas of gluons , quarks and antiquarks , which are seen as the effective degrees of freedom propagating in the plasma .
this assumption is actually common to all the so - called quasiparticle approaches @xcite .
however , thanks to the @xmath6-matrix formulation , the strongly - interacting regime can also be investigated here , in which bound states are expected to still survive above @xmath1 @xcite . the paper is organized as follows .
[ tmatsec ] is a summary of the approach used here and about which detailed explanations can be found in @xcite . in sec .
[ param ] , the model parameters are presented and discussed . in particular , the quasiparticle bare masses are extracted from the @xmath7 spectrum . in sec .
[ bs_qgp ] , the binary bound state spectrum above @xmath1 is computed and analysed . then , the eos of the deconfined phase at zero baryonic potential are studied for @xmath2 and @xmath3 in sec .
[ eos_qgp ] . to finish ,
an exploratory work at small baryonic potential is carried out in sec .
[ eos_qgp_mu ] .
all our eos are compared to recent lattice qcd ( lqcd ) ones .
our results are finally summarized in sec .
[ conclu ] .
the results of dashen , ma and bernstein @xcite establishing the grand potential of an interacting relativistic particle gas @xmath9 , expressed as an energy density , is given by ( in units where @xmath10 ) @xmath11\text{.}\ ] ] this equation is made of two parts .
the first term @xmath12 refers to the grand canonical potential of the free relativistic ( quasi)particles , while the second term accounts for interactions in the plasma .
this latter is made of a sum running on all the species , the number of particles included , and the quantum numbers necessary to fix a channel .
the vectors @xmath13 and @xmath14 contain the chemical potentials and the particle number of each species taking part in a given scattering process .
the set of all these channels is generically denoted @xmath15 . as usual , the chemical potential @xmath16 is the lagrange multiplier associated with the number of particles .
it is a measure for the density of particles . in relativistic models ,
the sign of @xmath16 is changed , passing from matter to antimatter . this parameter marks imbalance between matter and antimatter @xcite .
one can notice that the contribution of the bound and scattering states are decoupled .
the threshold @xmath17 is the summation on the masses of all the particles included in a given channel @xmath15 .
below @xmath17 , bound states appearing as pole in the @xmath18-matrix ( equivalently @xmath0-matrix ) are added as free additional species : @xmath19 is the grand canonical potential describing a free relativistic gas of the @xmath15-channel bound states . above @xmath17 ,
the scattering contribution is expressed as an integration depending on a trace , taken in the center - of - mass frame of the particles in the channel @xmath15 , and function of the @xmath18-matrix of the system .
@xmath18 is in particular a function of the total energy @xmath20 .
the symmetrizer @xmath21 enforces the pauli principle when a channel involving identical particles is considered , and the subscript @xmath22 means that only the connected scattering diagrams are taken into account .
@xmath23 is the modified bessel function of the second kind and @xmath24 where @xmath0 is the temperature .
the symbol @xmath25 denotes @xmath26 . by definition ,
@xmath18 is linked to off - shell @xmath0-matrix @xmath27 : @xmath28 where @xmath29 is the free hamiltonian of the system . as in @xcite , we will only focus on two - body channels .
so , a way to obtain @xmath27 is to solve the lippmann - schwinger equation , schematically given by @xmath30 with @xmath31 the free two - body propagator and @xmath32 the interaction potential .
it is worth mentioning that for three - body channels , faddeev equations should be used in order to eliminate the spurious solution of the lippmann - schwinger equation @xcite .
such considerations will be thus out of scope in this paper .
once eq .
( [ pot0 ] ) is computed , all thermodynamic observables can derived .
for example , the pressure is simply given by @xmath33 the sum @xmath34 appearing in ( [ pot0 ] ) explicitly reads @xmath35 , where only two particles are involved in the interaction process , @xmath36 is a possible isospin channel , @xmath37 is the color channel , and @xmath38 is the spin / helicity channel ( the labels @xmath39 or @xmath40 must be dropped off if the charge conjugation or the parity are not defined ) .
the normalized trace anomaly can also be computed by the following formula @xmath41 where @xmath42 is the stefan - boltzmann pressure .
although we give here some results about the trace anomaly , it is mentioned in @xcite that some improvements must be done in order to obtain a fully reliable estimation of this quantity .
assuming that the dominant scattering processes are the two - body ones , a key ingredient of the present approach is the two - body potential @xmath32 , encoding the interactions between the particles in the plasma . as in @xcite , @xmath32 is extracted from the static quenched su(3 ) lqcd free energy @xmath43 , between a @xmath44 pair in singlet representation at finite temperature @xcite , and then fitted with a cornell potential , screened thanks to the debye - hckel theory @xcite(see appendix b in @xcite ) .
note that unquenched lqcd results are also available in @xcite .
nevertheless , since these results are not significantly different from the quenched ones , the quenched potential is kept as basis of our computations , giving the accuracy expected in our work . from that ,
the internal energy @xmath45 is computed , @xmath46 , and considered as the interaction potential .
this choice is still a matter of debate .
nevertheless , it has given correct results in the ordinary ym case , as shown in @xcite .
moreover in sec [ qgp_3 ] , we will see that , according to our prescription for the quasiparticle masses , the internal energy is required to have a better agreement between our results and lqcd ones just above @xmath1 .
no relativistic corrections will be taken into account for light - quark interactions within this paper .
indeed , the quasiparticle quark masses used in our approach are large enough to assume static potentials at first approximation .
nevertheless , this task is left for future works .
moreover , all hyperfine interactions are neglected .
we can expect that they are non - dominant with respect to the spin - independent contributions , since these processes are assumed to depend on the inverse square of the effective mass . with this hypothesis , we also miss the diagonal annihilation contributions . finally , the casimir scaling is used to extract the leading - order gauge dependence of @xmath47 for @xmath48 , as proposed in sec .
ii in @xcite .
the casimir scaling means that potentials between colored sources are proportional to the eigenvalues of the quadratic casimir operator for their representation @xcite .
it is the simplest color dependence for two color sources : it has indeed the same form as the one for the one - gluon exchange process .
nevertheless , it is important to stress that the interaction considered within this paper contains other processes since it stems from a lqcd computation .
let us note that the annihilation mechanism , which does not respect the casimir scaling , is a contact interaction and is then vanishing for all non - s states .
moreover , it is worth mentioning that the casimir scaling seems very well respected between two static color sources in the @xmath49 sector @xcite .
computations in the @xmath50 sector show a situation which is slightly different : the casimir scaling seems partly violated ( at most 20% ) for short distances and temperatures near @xmath1 @xcite . nevertheless in this work , as in @xcite , we assume that the casimir scaling is satisfied .
the final form of the potential is thus the following one : @xmath51 , \label{used_pot}\ ] ] where @xmath52 and where @xmath53 is the quadratic casimir of the representation @xmath54 .
@xmath55 , adj , @xmath56 and @xmath57 stand respectively for the pair , adjoint , @xmath58- and @xmath59-particle representation . for instance , @xmath60 for the su(n ) gauge group ( the singlet representation is denoted by @xmath61 ) .
all the values taken by for the various color channels considered in this study are given in appendix a in @xcite .
let us note that the interaction can be attractive or repulsive .
the normalization of is given by @xmath62 , since @xmath47 is fitted on a singlet @xmath44 potential for su(3 ) .
we can also notice in that the long - distance behavior of the lattice potential @xmath63 , is subtracted .
indeed , this term is assimilated , as suggested in @xcite , as a thermal mass contribution for the quasiparticles .
moreover , it ensures the convergence of the scattering equation and the possibility to perform the fourier transform . when the quasiparticles are infinitely separated , the only remaining potential energy can be seen as a manifestation of the in - medium self - energy effects , @xmath64 .
we thus encode these effects as a mass shift @xmath65 to the bare " quasiparticle mass @xmath66 , by following the arguments exposed in @xcite : @xmath67 in order to get the thermal mass for any particles , the first - order color dependence is extracted in agreement with the hard - thermal - loop ( htl ) leading - order behavior @xcite : @xmath68 where the quantity @xmath69 is assumed to be color - independent . as @xmath47 is fitted on a singlet @xmath44 potential for @xmath70 , we have here @xmath71 for further details about the behavior of @xmath72 , one can refer to sec .
v in @xcite and to sec .
[ param ] in this paper . at this stage ,
one has to have in mind that chiral symmetry is not taken into account in our formalism .
comments about that issue will be given in the conclusions .
the lippman - schwinger equation leading to the on - shell @xmath0-matrix can be computed from as in @xcite : @xmath73 \left[1 \pm f_{p2}(\epsilon_2)\right ] , \nonumber\end{aligned}\ ] ] where @xmath20 is the energy in the center - of - mass frame , @xmath74 the asymptotic energy of the particle @xmath58 , and where the free two - body propagator is computed thanks to the blanckenbecler - sugar ( bbs ) reduction scheme .
its explicit form is given in appendix c in @xcite .
moreover , the in - medium effects , namely the bose - enhancement and the pauli - blocking are included following @xcite .
@xmath75 is thus the distribution function of the @xmath76-species : @xmath77 the @xmath78 stands for bosons while the @xmath79 for fermions , and @xmath16 is a possible chemical potential .
the sign choice in also depends on the nature of the particles : @xmath79 for bosons and @xmath78 for fermions .
let us note that the impact of these in - medium effects on our eos is very small .
therefore , the results obtained in @xcite remain valid .
concerning the interaction potential @xmath80 entering in , it is obtained by the fourier transform of the interaction extracted in lqcd .
since our potential has a spherical symmetry , we have @xmath81 where @xmath82 and @xmath83 is the angle between the momenta @xmath84 and @xmath85 . for channels given by ordinary @xmath86 states
, @xmath80 is obtained from @xmath87 where @xmath88 is the legendre polynomial of order @xmath89 . the spin @xmath18 is not indicated since our interaction is spin - independent . when at least one particle is transverse , the helicity formalism @xcite has to be used .
it is then very convenient to decompose a helicity state in the basis states @xmath86 in order to perform the computations . for a particular helicity state @xmath90
, it reads @xmath91 then , it can be shown that @xmath92 since our interaction is spin - independent .
all the helicity states needed for this study are listed in appendix b in @xcite .
the haftel - tabakin algorithm is a reliable procedure to solve the @xmath0-matrix problem @xcite .
the momentum integral is discretized within an appropriate quadrature , thus turning the integral equation into a matrix equation , namely @xmath93 , where schematically , @xmath94 and where @xmath95 denotes the integration weight .
the solution follows trivially by matrix inversion .
bound states are naturally poles below @xmath17 .
an interesting criterion for finding them is to use the determinant of the transition function @xmath96 ( referred to as the fredholm determinant ) since it vanishes at the bound state energies @xcite .
finally , once @xmath97 is known , the on - shell @xmath0-matrix is readily obtained as @xmath98 , with @xmath99 given by @xmath100
before fixing the parameters and applying the general formalism described in the previous section to the @xmath101 qgp , let us discuss some general assumptions that we have done within this model . in our approach
, there are different species of quasiparticles : the gluons ( @xmath102 ) , the light ( anti)quarks ( @xmath103 , resp .
@xmath104 ) , the strange ( anti)quarks ( @xmath105 , resp .
@xmath106 ) and the heavy ( anti)quarks ( @xmath22 and @xmath107 , resp . @xmath108 and @xmath109 ) . the ( anti)quarks are spin-1/2 particles belonging to fundamental ( conjugate ) representation of the gauge group . despite their nonvanishing mass ,
the gluons are transverse spin-1 bosons in the adjoint representation .
the gluon mass is dynamically generated by self - energy effects , which does not imply a drastic change of their nature .
it has been shown that the gluon must be considered as a transverse spin-1 boson to reproduce correctly the expected glueball spectra at @xmath7 @xcite . moreover ,
some lattice data support the presence of massive transverse modes only in a gluon plasma @xcite .
the two - body channels to be considered are @xmath110 , @xmath111 , @xmath112 , @xmath44 , @xmath113 and @xmath114 .
the lowest corresponding spin / helicity states are given in appendix b of @xcite , and the possible color channels can be found in appendix a of @xcite . within this study , we only focus on the su(3 ) gauge group .
let us examine the different possibilities of interactions : * as explained above , the interaction ( [ used_pot ] ) between two gluons or two quarks follows strictly the casimir scaling and neglects all hyperfine corrections , annihilation ones included . *
although this interaction is expected to take into account complicated exchanges ( since it stems from a lqcd calculation ) , it is interesting to look at the simplest possible feynman diagrams between two particles .
two gluons or two quarks can exchange a gluon , but the basic gluon - quark interaction is a quark exchange .
so the choice ( [ color_scaling ] ) for the color factor is questionable for this particular interaction . to correct
this point is beyond the scope of this work , but it is worth mentioning that the contributions of the @xmath113 and @xmath114 interactions is expected to be very weak in our model ( see sec .
[ eos_qgp ] ) .
* processes transforming a @xmath110 pair into a @xmath115 pair exist , but we have checked that mechanisms of order 1 are naturally suppressed since there is no overlap between @xmath110 and @xmath115 states @xcite .
as we neglect second order processes , as hyperfine interactions , we do not take into account transition between @xmath110 and @xmath115 pairs . in order to fix our parameters for starting the computations at finite temperature , and to check the validity of our model
, some pieces of information can be extracted from the @xmath7 bound state spectrum as in @xcite . in quenched su(3 ) lattice qcd , the potential between a static quark - antiquark pair at zero temperature is compatible with the funnel form @xmath117 where @xmath118 and @xmath119 gev@xmath120 ( standard values for the running coupling constant @xmath121 and the string tension @xmath122 at @xmath49 ) .
again , we neglect the contributions of annihilation processes . since the fourier transform of @xmath123 is not defined ( because of a non - zero asymptotic value )
, a string - breaking value @xmath124 , has to be introduced in order to make it convergent @xcite .
@xmath124 is thus seen as the energy above which a light quark - antiquark pair can be created from the vacuum and breaks the qcd string .
this scale is then subtracted and the potential effectively taken into account is @xmath125 , while @xmath126 is interpreted as an effective quark mass using the same arguments as those detailed in sec . [ properties ] . according to the color scaling ,
the potential describing the interactions between two color sources ( with representations @xmath54 and @xmath127 ) at zero temperature , is @xmath128 since @xmath129 .
the factor @xmath130 appears since the potential @xmath131 is fitted on a singlet @xmath44 pair for a su(3 ) gauge group . in this case , @xmath132 should rather be interpreted as the energy scale necessary to form two sources of color compatible with the existence of the two new color singlet pairs of particles created by the string breaking . as in @xcite , if @xmath66 is the bare mass of the particle , the @xmath49 mass @xmath133 , used to compute the bound state is then @xmath134 keeping the same structure as in .
the zero - temperature spectrum of the theory can be computed by solving with the potential . as mentioned in sec .
[ lseq ] , instead of looking at the pole of the @xmath0-matrix , the zeros of @xmath135 are computed in order to establish the bound state spectrum . the lightest glueball spectrum , namely the @xmath136 , @xmath137 and @xmath138 , has already been computed in @xcite . the parameters , @xmath139 and @xmath140 , were respectively fixed to 2 gev and 0.7 gev .
@xmath139 is in agreement with lattice data showing that the mass of the lightest gluelump is given by @xmath141 gev @xcite , while @xmath140 is an acceptable value for the zero - momentum limit of the gluon propagator at zero temperature in view of previous studies locating this mass typically between 500 and 700 mev ( see _ e.g. _ @xcite ) .
interesting reader can refer to @xcite for additional information about the @xmath7 glueball spectrum . within this paper ,
the stress is put on mesons with an orbital angular momentum @xmath142 or @xmath143 .
the allowed states with these quantum numbers are displayed in table [ tab_qqbar ] .
as it can be seen , several @xmath38 states are associated to a same @xmath89 .
since the potential does not depend on other quantum numbers , all these states are degenerate within our approach .
.[tab_qqbar ] @xmath38 states allowed for @xmath44 with @xmath142 or @xmath143 at @xmath7 .
the parity of the state @xmath40 , is given by @xmath144 while the charge conjugaison @xmath39 , is @xmath145 . [ cols="^,^,^,^",options="header " , ] finally , let us add that , unlike in the glueball case , the @xmath7 meson mass depends on the gauge group since @xmath146 depends on it ( see appendix a in @xcite ) . within our approach , such study is not difficult to carry out .
the meson mass dependence in function of the gauge group is not studied here , since the principal interest of the @xmath0-matrix computations at @xmath7 is to extract and to check the parameters we will use at @xmath147 . in this regard , let us note that only the quasiparticle bare masses will enter in our computations at @xmath147 and not the string breaking @xmath124 .
what can be said however about the gauge - group dependence is that , in the case of su(@xmath148 ) , the meson masses are of order 1 as expected , see conclusions for more comments about this .
the last global parameter that has to be fixed within our approach is the critical temperature of deconfinement @xmath1 .
as we fit our interactions on lattice calculations , the definition of @xmath1 comes from these approaches : the color averaged as well as the singlet free energy of a quark - antiquark pair will tend towards finite nonzero values for all temperatures @xmath48 ( and diverge below @xmath1 ) . in @xcite ,
a value of 0.3 gev was used since the focus was only on the gluon sector . here , in order to stay coherent with current lattice data , @xmath1 is moved to 0.15 gev .
this change naturally modifies the thermodynamics in the gluon sector that was established in @xcite .
let us discuss this point .
first of all , let us set @xmath149 .
the two - body interaction potential between particles only depends on @xmath150 as it can be explicitly shown from its expression given in appendix b of @xcite .
therefore , it is the same for the gluon thermal mass @xmath151 , according to .
moreover , since the effect of the bose - enhancement can be considered as negligible , it can also be assumed that the @xmath0-matrix @xmath152 , is only a function of @xmath150 . in fig .
[ comparetemp ] , the @xmath1-impact is analyzed for the pure gauge eos with @xmath153 0.15 and 0.30 gev , all other parameters remaining fixed . the way to compute these eos is given in @xcite and will be recalled in sec .
[ gen_eos ] .
as it can be noticed , the general behavior is not the same and the normalized pressure seems to increase when @xmath1 increases .
it is especially worth remarking that , if we only consider the contribution of the quasiparticle ideal gas , @xmath1 has an impact on the thermodynamics .
indeed , the pressure depends on the ratio @xmath154 , depending on @xmath1 for a fixed @xmath150 . , and trace anomaly @xmath155 ( without bound states ) versus temperature in units of @xmath1 ( with @xmath156 and 0.3 gev).,title="fig:",scaledwidth=49.0% ] , and trace anomaly @xmath155 ( without bound states ) versus temperature in units of @xmath1 ( with @xmath156 and 0.3 gev).,title="fig:",scaledwidth=49.0% ] the behavior of the normalized trace anomaly ( without bound states ) is also presented in fig . [ comparetemp ] .
the @xmath1-dependence is not very easy to predict because the trace anomaly depends on the slope of the associated pressure curve , and small variations can generate a drastic change . in fig .
[ comparetemp ] , we can indeed observe that the behavior around @xmath1 is extremely different for @xmath156 and 0.3 gev .
the peak structure is lost for @xmath156 gev .
this could be due to two reasons : the total change of the free part structure and the small impact of the interactions in comparison with the eos obtained in @xcite , as it is shown in fig .
[ comparetemp150 ] . versus temperature in units of @xmath1 ( with @xmath156 gev ) ,
compared to the free part contribution .
( right ) normalized pure - gauge trace anomaly @xmath155 versus temperature in units of @xmath1 ( with @xmath156 gev ) , compared to the free part contribution.,title="fig:",scaledwidth=49.0% ] versus temperature in units of @xmath1 ( with @xmath156 gev ) , compared to the free part contribution .
( right ) normalized pure - gauge trace anomaly @xmath155 versus temperature in units of @xmath1 ( with @xmath156 gev ) , compared to the free part contribution.,title="fig:",scaledwidth=49.0% ]
the existence or not of bound states in the deconfined phase is not forbidden in principle , especially around @xmath1 where interactions are expected strong enough to bind two or more particles @xcite . since the operator @xmath157 is negative for several color channels , the finite - temperature spectrum of qcd above @xmath1 can be computed by solving with the potential .
the thermal masses of the quasiparticles are given by with @xmath66 extracted from the @xmath7 bound state spectrum ( see sec .
[ param_t0 ] ) . within our formalism
, the channels in which bound states are favored at most should contain a @xmath18-wave component to avoid the centrifugal barrier and should have a symmetry that allows the state to be in a color singlet , the color channel in which the interactions are maximally attractive . &
@xmath158 & @xmath159 + & singlet & as + @xmath160 & @xmath161 & & + 1.05 & 1.67 & 1.51 & 1.67 + 1.10 & 1.28 & - & - + & @xmath158 & @xmath159 + & singlet & as + @xmath160 & @xmath162 & & + 1.05 & 1.71 & 1.54 & 1.71 + 1.10 & 1.43 & -&- + & @xmath158 & @xmath159 + & singlet & as + @xmath160 & @xmath163 & & + 1.05 & 1.76 & 1.57 & 1.74 + 1.10 & 1.38 & 1.38 & - + 1.15 & 1.23 & - & + in the @xmath110 case there are two such states : the @xmath136 and @xmath138 ones , in color singlet , correspond to the scalar and tensor glueballs respectively .
we have observed in @xcite that both the scalar and tensor glueball masses at @xmath164 @xmath1 were compatible with the zero - temperature ones .
moreover , the scalar glueball exists as a bound state up to 1.25 @xmath1 while the tensor one is bound up to 1.15 @xmath1 . note that in @xcite , the impact of the bose - enhancement were not considered in the @xmath0-matrix .
therefore , the value of @xmath1 should modify the masses of the bound states .
nevertheless , it has been numerically checked that the data shown in @xcite differ from the ones containing in - medium effects only with a relative error of the order of 2% .
this is the reason why they are not presented again here .
concerning the light and strange mesonic sector , the dissolution inside the plasma is much more rapid .
indeed , the most attractive channel , the @xmath142 in singlet , is the only one that survives above @xmath1 .
moreover , mesons quickly dissolve inside the plasma as it can be observed in table [ tab : all1 ] , and the meson masses are not compatible with the @xmath7 one as in the @xmath110 case .
the same assertion can also be drawn for the @xmath111 and @xmath112 sector . indeed ,
since there is no more confinement ( only singlet representation ) in the qgp , such states could exist above @xmath1 ; the most attractive one being the @xmath142 in the antisymmetric representation ( as ) .
nevertheless , they also rapidly disappear just above @xmath1 , as shown in table [ tab : all1 ] .
such rapid dissolution in comparison with the @xmath110 case can be understood by the fact that the quark quasiparticle mass is lighter than the gluon one and that the @xmath165 is more than two times the maximum magnitude of @xmath166 in the quark sector .
indeed , @xmath167 while @xmath168 and @xmath169 for a su(3 ) gauge group ( see appendix a of @xcite ) . note that similar comments can also be done about the @xmath113 and @xmath114 sector , leading to a quick melting of these bound states inside the plasma ( see table [ qg_diss ] ) .
& @xmath158 & @xmath159 & @xmath170 & @xmath171 + & singlet & as & singlet & as + @xmath160 & @xmath172 & & & & + 1.05 & 3.56 & 3.14 & 3.44 & 3.48 & - + 1.10 & 3.39 & 3.20 & 3.36 & - & + 1.15 & 3.33 & 3.22 & - & & + 1.25 & 3.28 & 3.26 & & & + 1.35 & 3.26 & - & & & + concerning the heavy quark sector , quarkonia have already been studied within a similar @xmath0-matrix approach as the one proposed here @xcite .
the main differences are the inclusion of a relativistic correction to the potential in @xcite and the way of implementing the quasiparticle masses . within this paper , the procedure to determine the quasiparticle masses
is applied and allows one to compute systematically a large panel of binary bound states made of different quark flavors .
& @xmath158 & @xmath159 & @xmath170 & @xmath171 + & singlet & as & singlet & as + @xmath160 & @xmath173 & & & & + 1.05 & 10.2 & 9.35 & 9.75 & 9.60 & 9.90 + 1.15 & 9.94 & 9.62 & 9.87 & 9.79 & 9.94 + 1.20 & 9.93 & 9.71 & 9.86 & 9.92 & - + 1.25 & 9.93 & 9.70 & 9.90 & - & + 1.30 & 9.92 & 9.75 & 9.90 & & + 1.50 & 9.92 & 9.79 & - & & + 2.00 & 9.91 & 9.88 & & & + 2.40 & 9.91 & - & & & + our study for the heavy quarkonia is displayed in tables [ quarkonia_c ] and [ quarkonia_b ] . around @xmath1 , the @xmath174 and @xmath175 masses are compatible with the @xmath7 spectrum , unlike for the light and strange mesons .
moreover , they significantly survive above @xmath1 , even if the dissociation temperatures are lower than the ones found in @xcite .
@xmath111 and @xmath112 states can also be formed with the medium but they dissolve more rapidly than the associated quarkonia , due their weaker interaction potential .
an analyse for the @xmath176 and @xmath177 mesons has also been carried out as well as for the @xmath113 and @xmath114 states for all the quark flavor considered here .
the different temperatures of dissociation are displayed in tables [ meson_diss ] , [ qq_diss ] and [ qg_diss ] .
we can notice that the more the quasiparticles considered in the binary state are heavy , the more it survives significantly above @xmath1 .
& light & strange & charm & beauty + light & 1.10 @xmath178 0.05 & 1.10 @xmath178 0.05 & 1.15 @xmath178 0.05 & 1.20 @xmath178 0.05 + strange & & 1.15 @xmath178 0.05 & 1.15 @xmath178 0.05 & 1.20 @xmath178 0.05 + charm & & & 1.35 @xmath178 0.05 & 1.60 @xmath178 0.1 + beauty & & & & 2.4 @xmath178 0.1 + & light & strange & charm & beauty + light & 1.10 @xmath178 0.05 & 1.10 @xmath178 0.05 & 1.10 @xmath178 0.05 & 1.10 @xmath178 0.05 + strange & & 1.10 @xmath178 0.05 & 1.10 @xmath178 0.05 & 1.10 @xmath178 0.05 + charm & & & 1.15 @xmath178 0.05 & 1.25 @xmath178 0.05 + beauty & & & & 1.50 @xmath178 0.05 + light & strange & charm & beauty + 1.10 @xmath178 0.05 & 1.10 @xmath178 0.05 & 1.20 @xmath178 0.05 & 1.25 @xmath178 0.05 + in @xcite , it is found that radially excited states are unlikely to survive above @xmath1 since they tend to melt below the phase transition because of string breaking effects at finite temperature . in our case , such states can be found but not in all channels .
however , they quickly dissolve within the medium .
now that the bound - state sector is analyzed , it is possible to compute explicitly the eos and so , to study the qgp thermodynamics . in what follows
, the heavy quark states will be not included in our eos .
indeed , their contributions to the bound - state and scattering parts of the grand canonical potential are expected to be small because of their large bound - state masses .
some preliminary lattice results about the influence of charm quarks on the eos can be found in @xcite .
it appears actually that charm quarks bring a significant contribution to the trace anomaly above @xmath1 .
however , a technical problem of the present approach is that discontinuity appear in the trace anomaly when bound states melt @xcite .
this problem is especially apparent when heavy quarks are involved .
hence , including heavy flavors in our computations would lead to results that are probably not reliable , and we prefer not to consider them .
let us thus particularize to a qgp with @xmath101 .
as in @xcite , a two - body restriction is used : the considered interactions are @xmath110 , @xmath111 , @xmath112 , @xmath44 , @xmath113 and @xmath114 , in different colour and @xmath180 channels .
therefore , the first term in ( [ pot0 ] ) , _ i.e. _ the free relativistic gas is given by @xmath181 where the gluons have a mass @xmath182 , and the ( anti)quarks @xmath183 ( @xmath184 ) a mass @xmath185 , given by the prescription with the @xmath66 value extracted from the @xmath7 spectrum .
@xmath186 is the chemical potential of the considered quark flavor .
they are set to zero within this section .
so , no asymmetry between quarks and antiquarks is taken into account within the qgp .
the particle degrees of freedom are the following .
the gluon is a transverse spin-1 ( so , two spin projections ) boson lying in the adjoint representation of the gauge group , while the quark ( resp .
antiquark ) , existing in @xmath187 different flavors , is a spin-1/2 fermion belonging in the fundamental ( resp .
conjugate ) gauge - group representation .
the grand canonical potential per degree of freedom associated to a bosonic species @xmath188 , and to a fermionic species @xmath189 , with mass @xmath190 are given by @xmath191 for later convenience , the thermodynamic quantities will be normalized to the stefan - boltzmann pressure , which is defined as @xmath192 and reads in this case @xmath193\text{.}\ ] ] as already mentioned , the second term of stands for the interactions .
the sum @xmath34 now explicitly reads @xmath194 , where @xmath195 , @xmath196 , @xmath197 are respectively the number of gluons , quarks and antiquarks involved in the interaction process .
attractive interactions can lead to the formation of bound states with masses @xmath198 ( see sec .
[ bs_qgp ] ) .
they contribute also to the grand potential as new species via the formula @xmath199 all @xmath200 and color channels leading to bound states are included in this summation .
concerning the scattering term , a tedious calculation ( explained in @xcite ) , leads to the following result , @xmath201 \right . \\ & & + \left .
\displaystyle\frac{1}{16 \pi^2 } \displaystyle\int_{m_1 + m_2}^\infty d\epsilon \ , \epsilon^2 \ , \omega(\epsilon)^2 \lambda(\epsilon)^2 k_2(\beta\epsilon ) \left[\left({\rm re } t_{{\cal c } , j^p}(\epsilon ; \omega(\epsilon ) , \omega(\epsilon))\right ) ' { \rm i m } t_{{\cal c } , j^p}(\epsilon ; \omega(\epsilon ) , \omega(\epsilon ) ) \right ] \right)\text { , } \nonumber \end{aligned}\ ] ] where @xmath202 and @xmath203 are given by @xmath204 and where @xmath205 is the on - shell @xmath206-matrix .
note that a isospin number @xmath36 has to be taken into account when one deals with @xmath207 and @xmath208 quarks since they have the same mass in our approach .
this isospin number enters in the summation @xmath209 as a @xmath210 factor .
finally , the grand canonical potential ( reduced to two - body interactions ) is summarized by the following formula @xmath211 for obvious numerical reasons , the summation over the number of particles is not the only one that must be restricted .
all possible color channels are included , but all the possible @xmath180 channels contributing to @xmath212 can not be included since their number is infinite .
so , a reliable criteria to select the most significant ones has to be established .
the basic idea , already proposed in @xcite , is that only states with low @xmath89 are included since they are the most likely to contribute significantly to a total mean cross section @xmath213 .
are only retained , the channels for which the value of @xmath213 is at least 25% of the value @xmath213 for the channel with the lowest value of @xmath214 ( see appendix d of @xcite for further explanations ) . in the present case
, this criterion implies that only the following @xmath180 channels are included . * for @xmath110 channels : the @xmath136 , @xmath137 , @xmath138 and @xmath215 states ; * for @xmath111 , @xmath112 and @xmath44 channels : all the @xmath216 ones with @xmath142 or @xmath143 ; * for the @xmath113 and @xmath114 channels : all the states with @xmath217 ( see appendix b in @xcite ) .
now that the number of @xmath216 channels for each two - body interactions is fixed , the eos can be computed . in fig .
[ qgp2_pressure ] , the normalized pressure is shown for a qgp with two light - quark flavor included .
as it can be noticed , interactions do not practically contribute : the major part is given by the free gas .
globally , the weakness of interactions can be interpreted in the same way as what it is observed for the ym plasma in sec .
[ param_tc ] . indeed , when the critical temperature decreases , the interactions seem to become smaller and smaller .
this behaviour is driven by the bessel functions entering in the definition of the scattering part .
moreover , it is important to notice that the integration range in this term formally starts at @xmath219 .
this value is large in comparison to the values at which the bessel functions is significantly non - zero .
a change of the thermal mass prescription could thus impact the contributions of the scattering part .
this work is left for further developments of our approach . versus temperature in units of @xmath1 ,
compared to the free part , bound state and scattering contribution .
( right ) different scattering contributions to the normalized pressure @xmath220 versus temperature in units of @xmath1 .
@xmath156 gev.,title="fig:",scaledwidth=49.0% ] versus temperature in units of @xmath1 , compared to the free part , bound state and scattering contribution .
( right ) different scattering contributions to the normalized pressure @xmath220 versus temperature in units of @xmath1 .
@xmath156 gev.,title="fig:",scaledwidth=49.0% ] in the right panel of fig .
[ qgp2_pressure ] , the different scattering contributions are separated . without surprise ,
the @xmath44 , @xmath113 and @xmath114 channels asymptotically tend to zero .
indeed , it has been shown in @xcite that the interactions between two different species vanish within the born approximation , because of an identity relating the color factors : @xmath221 . concerning the @xmath111 and @xmath112 channels , they generate a global increase of the normalized pressure while it is the contrary for the @xmath110 sector .
not only these two effects are weak but in addition , they contribute in opposite directions , leading to a global suppression of the two - body interactions in average .
( without bound states ) versus temperature in units of @xmath1 , compared to the free part contribution .
@xmath156 gev.,scaledwidth=49.0% ] in fig .
[ qgp2_anom ] , we display the normalized trace anomaly ( without bound states ) compared to the free gas part .
a peak structure is here exhibited even in the free gas contribution .
therefore , it is different from the ym sector where the interactions create the peak .
the nature of this latter is really difficult to establish since few variations of the pressure can drastically change the shape of the trace anomaly .
the main conclusion that seems to emerge from our approach ( looking at the normalized trace anomaly as well as at the normalized pressure ) is that the leading behavior of the qgp is driven by gluon and ( anti)quark degrees of freedom that interact weakly .
nevertheless , it does not mean that the interactions have no impact on the eos .
indeed , the particle thermal mass is extracted from it , leading to a self - energy contribution for the particle ( see sec .
[ qgp_3 ] ) . a similar analyse as the one proposed in the previous subsection can be applied in the 2 + 1 qgp case . since
similar results and features can be deduced from it , we will not repeat it again and focus more on the comparisons between our model and lqcd extracted from @xcite .
indeed , lqcd collaborations have recently reached the physical quark masses in their computations of the eos , making their results more and more reliable for comparisons . versus temperature in units of @xmath1 ,
compared to lqcd data from @xcite .
( right ) normalized trace anomaly @xmath155 versus temperature in units of @xmath1 , compared to lqcd data from @xcite .
@xmath156 gev.,title="fig:",scaledwidth=49.0% ] versus temperature in units of @xmath1 , compared to lqcd data from @xcite .
( right ) normalized trace anomaly @xmath155 versus temperature in units of @xmath1 , compared to lqcd data from @xcite .
@xmath156 gev.,title="fig:",scaledwidth=49.0% ] as we can observe in fig .
[ qgp21 ] , our data are qualitatively in agreement with lqcd ones .
the lqcd normalized pressure is slightly overestimated as well as the asymptotic behavior of the normalized trace anomaly . on the other hand ,
the peak structure of the trace anomaly is very different of the lqcd one .
as already mentioned , this latter is really difficult to obtain due to several reasons in our approach : problems in the inclusion of the bound state @xcite , reliability of the quark masses and restriction to two - body interactions . even in lqcd ,
different collaborations find different quantitative behaviors for the trace anomaly peak @xcite .
the possible discrepancies can arise from the choice of the fermionic lattice action , the lattice spacing , the considered quark masses , the extrapolation to the continuum limit ... only , a good agreement in the behavior of the decreasing tail is reached by the different lqcd groups , according to @xcite .
nevertheless , it is worth mentioning that the disagreements observed in the quantitative value of the peak structure in various lqcd results ( mainly due to a computation with no physical quark masses ) seem to reduce , and the shape of the lqcd trace anomaly tends to the one depicted in fig .
[ qgp21 ] , and firstly given by the bmw collaboration @xcite
. therefore , except for the normalized trace anomaly peak structure ( for which a more appropriate treatment of the bound - state inclusion is needed and could change significantly its structure ) , our data are in correct agreement with lqcd ones .
as discussed in the previous subsection , this agreement seems to be reached by only including a quasiparticle thermal mass : the contributions of the two - body interacting channels are minor .
nevertheless , it is worth insisting on the fact that the thermal mass effects are extracted from the two - body lqcd interaction potential within our model .
so , the chosen two - body interactions are not useless to understand the behavior of the qgp around @xmath1 . indeed , if we change the potential , the free gas contribution is modified since the quasiparticle thermal masses depend on it .
this leads to a completely different behavior of the eos as seen in fig .
[ uvsf ] in which the potential is now chosen to be the free energy .
we can especially notice in fig .
[ uvsf ] that a better agreement between our model and lqcd normalized pressure is reached around @xmath1 thanks to the internal energy while the discrepancy between the two curves decrease when the temperature increases . versus temperature in units of @xmath1 .
( right ) normalized trace anomaly @xmath155 ( without bound states ) versus temperature in units of @xmath1 .
in the two figures , the gray ( black ) line is the free part contribution of @xmath212 computed with the free ( internal ) energy .
@xmath156 gev .
, title="fig:",scaledwidth=49.0% ] versus temperature in units of @xmath1 .
( right ) normalized trace anomaly @xmath155 ( without bound states ) versus temperature in units of @xmath1 .
in the two figures , the gray ( black ) line is the free part contribution of @xmath212 computed with the free ( internal ) energy .
@xmath156 gev . ,
title="fig:",scaledwidth=49.0% ] finally , let us compare in fig [ qgp21vsqgp2 ] the normalized pressure and trace anomaly for a qgp with @xmath2 and @xmath3 ( @xmath153 0.15 gev ) to the ones of the ym plasma ( @xmath153 0.3 gev ) .
we can notice that the normalized pressure curves are almost superimposed and that the decreasing trend of the trace anomaly is nearly the same in all the considered theories .
the maximum of the deviation between these curves is around @xmath223 , at the localisation of the trace anomaly peak .
it is nevertheless important to remember that the critical temperature and the normalization are not the same in all the eos ( see ) .
however , within these units , a universality at large temperature ( @xmath224 ) seems to emerge . versus temperature in units of @xmath1 .
( right ) normalized trace anomaly @xmath155 ( without bound states ) versus temperature in units of @xmath1 .
@xmath156 gev for qgp and @xmath225 gev for ym .
, title="fig:",scaledwidth=49.0% ] versus temperature in units of @xmath1 .
( right ) normalized trace anomaly @xmath155 ( without bound states ) versus temperature in units of @xmath1 .
@xmath156 gev for qgp and @xmath225 gev for ym .
, title="fig:",scaledwidth=49.0% ]
now that the eos for the qgp are computed and favourably compared with lqcd , we can investigate the non - zero baryonic regime .
this latter deserves a lot of interests , especially in the area of the neutron star physics . indeed ,
since pioneering works @xcite about the existence of a deconfined phase in qcd , it was assumed that the core of the heaviest neutron star should be probably filled by a medium with a high nuclear density and in which the significant degrees of freedom should be the quarks .
therefore , getting the qcd eos at finite @xmath16 could shed some light in this field . up to now
, this task still remain difficult from first qcd principles .
remember that even in lqcd some conceptual troubles appear ( cfr .
sign problem ) and only perturbations around @xmath226 are meaningful .
therefore , it seems appealing to check whether or not quasiparticle approaches could help .
unfortunately at the present stage , some problems also appear in our formalism .
the main reasons are the following .
first , the dashen ma and bernstein formalism that we have used to compute the eos is based on a virial expansion in terms of @xmath227
. we are thus limited by construction to small baryonic potential .
indeed , increasing the baryonic potential is the same as increasing the density of particles : the many - body interactions are more and more likely to contribute .
so , the reduction to two - body interactions becomes a poor approximation _ a priori _ and some problems , other than a careful computation of all the channels , arise .
let us mention for instance , the absence of a helicity formalism for many - body systems in a potential approach , and the necessity to resort to faddeev and higher equations for more than two - body interactions .
moreover , when the density of particles increases , the notion of quasiparticle becomes more and more questionable .
another peculiar problem is the building of a coherent interaction in presence of baryonic potential . already at two bodies ,
no lqcd data are available to our knowledge .
it is not only important to define the potential between particles but also the quasiparticle mass , which seems to rule the main behavior of the eos at @xmath226 .
a way to circumvent this problem could be to use the htl expressions for the particle thermal mass , but it was not the bias adopted within this study .
indeed , the actual shape of our thermal masses are not the ones extracted from htl . for all these reasons , the study that follows will be only limited to small baryonic potentials .
we will thus keep the restriction to two - body interactions which can make sense in such a @xmath16-range .
moreover , the interaction potential and the quasiparticle thermal masses are the same as the ones used up to now , without the inclusion of the baryonic potential . of course , the obtained results must be considered as preliminary and are just intended to draw a general tendency .
the baryonic potential enters at two levels in our computations : in the @xmath0-matrices because of the in - medium effects , and in all the eos contributions as multiplicative factors .
fortunately as for the @xmath1-impact , it seems that the @xmath16-dependence on the @xmath0-matrix calculations is negligible ( see fig .
[ mu_impact ] ) .
therefore , these latter do not have to be recomputed at each @xmath16 , which drastically reduces the computational time .
-matrix ( @xmath44 in @xmath142 , here given as example ) in function of the energy for @xmath0 = 0.105 gev ( left ) , @xmath0 = 0.150 gev ( middle ) and @xmath0 = 0.300 gev ( right ) at different @xmath16 ( mev ) with @xmath1 = 0.15 gev .
, title="fig:",scaledwidth=32.0% ] -matrix ( @xmath44 in @xmath142 , here given as example ) in function of the energy for @xmath0 = 0.105 gev ( left ) , @xmath0 = 0.150 gev ( middle ) and @xmath0 = 0.300 gev ( right ) at different @xmath16 ( mev ) with @xmath1 = 0.15 gev .
, title="fig:",scaledwidth=32.0% ] -matrix ( @xmath44 in @xmath142 , here given as example ) in function of the energy for @xmath0 = 0.105 gev ( left ) , @xmath0 = 0.150 gev ( middle ) and @xmath0 = 0.300 gev ( right ) at different @xmath16 ( mev ) with @xmath1 = 0.15 gev .
, title="fig:",scaledwidth=32.0% ] in fig .
[ pressure_mu ] , we have plotted the normalized pressure and trace anomaly at different @xmath16 for a qgp with two light quarks . the normalization is given by , that is to say at @xmath228 and @xmath226 . naturally , the gluon chemical potential is zero and the quark one is such that @xmath229 .
we can see in this figure that the normalized pressure increases with @xmath16 .
this pressure is especially driven by the increase of the free quark gas contribution given in fig . [ interaction_mu1 ] .
indeed , as in the @xmath226 case , the leading contributions to the normalized pressure are the free part ones since the impact of the interactions is small as observed in fig .
[ interaction_mu2 ] .
moreover , the decrease of the free antiquark gas contribution is slower than the increase of the free quark gas one , explaining the total increasing behavior of the normalized pressure . versus temperature in units of @xmath1 at different @xmath16 ( mev ) for a qgp with two light quarks .
( right ) normalized trace anomaly @xmath155 ( without bound states ) versus temperature in units of @xmath1 at different @xmath16 ( mev ) for a qgp with two light quarks .
@xmath156 gev.,title="fig:",scaledwidth=49.0% ] versus temperature in units of @xmath1 at different @xmath16 ( mev ) for a qgp with two light quarks .
( right ) normalized trace anomaly @xmath155 ( without bound states ) versus temperature in units of @xmath1 at different @xmath16 ( mev ) for a qgp with two light quarks .
@xmath156 gev.,title="fig:",scaledwidth=49.0% ] concerning the normalized trace anomaly , it is much more difficult to understand the @xmath16-dependence .
the only assertion that we can do is that the trace anomaly peak becomes higher and higher with the increase of @xmath16 .
moreover , we can notice that the convergence to zero is faster with large @xmath16 . versus temperature in units of @xmath1 at different @xmath16 ( mev ) for a qgp with two light quarks .
( right ) free antiquark gaz contribution to the total normalized pressure @xmath220 versus temperature in units of @xmath1 at different @xmath16 ( mev ) for a qgp with two light quarks .
@xmath156 gev.,title="fig:",scaledwidth=49.0% ] versus temperature in units of @xmath1 at different @xmath16 ( mev ) for a qgp with two light quarks .
( right ) free antiquark gaz contribution to the total normalized pressure @xmath220 versus temperature in units of @xmath1 at different @xmath16 ( mev ) for a qgp with two light quarks .
@xmath156 gev.,title="fig:",scaledwidth=49.0% ] as already mentioned , the scattering contributions are small . nevertheless in fig .
[ interaction_mu2 ] , we can observe a significant dependence in terms of @xmath16 .
the @xmath111 and @xmath113 scattering contributions obviously increase with @xmath16 , respectively as @xmath230 and @xmath231 , while the @xmath112 and @xmath114 ones go in opposite way .
however , as in the free gas case , the increase is higher than the decrease , leading _ in fine _ to a more important contribution of the scattering parts to the total normalized pressure . to be complete ,
the @xmath44 scattering contribution is stable since there is no @xmath16-dependence at the level of the eos : indeed , we have @xmath232 and just a very weak dependence on @xmath16 appears in the @xmath0-matrix . moreover
, since the @xmath110 sector is independent of @xmath16 , increasing @xmath16 means increasing the impact of the quark sector within the qgp .
-scattering contribution to the total normalized pressure @xmath220 .
( right , top ) same for @xmath112 .
( left , middle ) same for @xmath113 .
( right , middle ) same for @xmath114 .
( bottom ) same for @xmath115 .
all the scattering contributions are presented versus temperature in units of @xmath1 , with @xmath156 gev , at different @xmath16 ( mev ) for a qgp with two light quarks.,title="fig:",scaledwidth=40.0% ] -scattering contribution to the total normalized pressure @xmath220 .
( right , top ) same for @xmath112 .
( left , middle ) same for @xmath113 .
( right , middle ) same for @xmath114 .
( bottom ) same for @xmath115 .
all the scattering contributions are presented versus temperature in units of @xmath1 , with @xmath156 gev , at different @xmath16 ( mev ) for a qgp with two light quarks.,title="fig:",scaledwidth=40.0% ] -scattering contribution to the total normalized pressure @xmath220 .
( right , top ) same for @xmath112 .
( left , middle ) same for @xmath113 .
( right , middle ) same for @xmath114 .
( bottom ) same for @xmath115 .
all the scattering contributions are presented versus temperature in units of @xmath1 , with @xmath156 gev , at different @xmath16 ( mev ) for a qgp with two light quarks.,title="fig:",scaledwidth=40.0% ] -scattering contribution to the total normalized pressure @xmath220 .
( right , top ) same for @xmath112 .
( left , middle ) same for @xmath113 .
( right , middle ) same for @xmath114 .
( bottom ) same for @xmath115 .
all the scattering contributions are presented versus temperature in units of @xmath1 , with @xmath156 gev , at different @xmath16 ( mev ) for a qgp with two light quarks.,title="fig:",scaledwidth=40.0% ] -scattering contribution to the total normalized pressure @xmath220 .
( right , top ) same for @xmath112 .
( left , middle ) same for @xmath113 .
( right , middle ) same for @xmath114 .
( bottom ) same for @xmath115 .
all the scattering contributions are presented versus temperature in units of @xmath1 , with @xmath156 gev , at different @xmath16 ( mev ) for a qgp with two light quarks.,title="fig:",scaledwidth=40.0% ] finally , we close this study by comparing our preliminary results to the lqcd ones given by @xcite . within this paper
, they deal with a qgp with @xmath3 and with a small baryonic potential @xmath233 .
each flavor of quarks is considered to carry one third of @xmath233 .
therefore , we analyse the 2 + 1 qgp with @xmath234 . as for the @xmath226 case , we sightly overestimate the normalized pressure and we miss the peak of the normalized trace anomaly .
therefore , it seems that these differences have mainly the same origin as at @xmath226 and our extrapolations at small @xmath16 is compatible with lqcd .
compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
( right ) normalized trace anomaly @xmath155 compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
@xmath156 gev.,title="fig:",scaledwidth=40.0% ] compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
( right ) normalized trace anomaly @xmath155 compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
@xmath156 gev.,title="fig:",scaledwidth=40.0% ] compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
( right ) normalized trace anomaly @xmath155 compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
@xmath156 gev.,title="fig:",scaledwidth=40.0% ] compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
( right ) normalized trace anomaly @xmath155 compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
@xmath156 gev.,title="fig:",scaledwidth=40.0% ] compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
( right ) normalized trace anomaly @xmath155 compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
@xmath156 gev.,title="fig:",scaledwidth=40.0% ] compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
( right ) normalized trace anomaly @xmath155 compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
@xmath156 gev.,title="fig:",scaledwidth=40.0% ] compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
( right ) normalized trace anomaly @xmath155 compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
@xmath156 gev.,title="fig:",scaledwidth=40.0% ] compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
( right ) normalized trace anomaly @xmath155 compared to lattice qcd @xcite versus temperature in units of @xmath1 at different @xmath233 ( mev ) for a qgp with @xmath3 .
@xmath156 gev.,title="fig:",scaledwidth=40.0% ]
the present work is part of a program aiming at studying the thermodynamic properties of gauge theories in the deconfined phase .
the interested reader may read @xcite and @xcite for pure ym and susy ym theories , while this paper is devoted to the realistic " quark - gluon plasma .
the framework developed is based on a @xmath0-matrix formulation of statistical mechanics , in which the thermal masses and two - body interactions are derived from the static potential between fundamental color source computed in quenched lattice qcd @xcite .
apart from the potential , the only remaining parameters are the value of @xmath1 and the bare quark masses .
these masses are fitted on the meson spectrum at zero temperature .
the main assumption underlying our model is actually that a quasiparticle picture of deconfined matter just above deconfinement is relevant . although it is not a rigorous proof , the nice agreement between our computed equations of state an the recent lattice data of refs .
@xcite can be seen as an _ a posteriori _ validation of our framework .
* both the free energy or the internal energy could be used as potential terms in our model .
it appears that , keeping the same procedure and the same values for the parameters , only the internal energy is able to generate an equation of state which has the qualitative features of the lattice equation of state .
the internal energy thus appears as the most relevant potential within in our framework and leads to a good agreement with the lattice equation of state .
note that this problem is far from being elucidated , see for example the recent work @xcite , where the opposite conclusion is reached . * between 1 and 2 @xmath1 , color interactions are strong enough to create mesons ,
_ i.e. _ a quark - antiquark bound state in a color singlet .
mesons made of one or two light quarks are almost all dissociated in @xmath1 .
only mesons made of two heavy quarks ( @xmath22 , @xmath107 ) are bound enough to survive in the range @xmath235 @xmath1 .
although we use a @xmath0-matrix formulation as well , our parameters have different values of that used in @xcite where the main goal was to reproduce mesonic correlators computed on the lattice and not the equation of state . in this last work ,
the @xmath174 meson is bound up to 3 @xmath1 and the @xmath175 meson is bound up to 3.5 @xmath1 , thus at much higher @xmath0 than what we find .
it is worth recalling that we are able to compute @xmath0-matrices in channels where the quark and the antiquark have different masses , which was not considered in @xcite . *
although strong in the color singlet channel , the contribution of two - body interactions to the equation of state is weak with respect to the free - gas part .
this is partly due to a cancellation between attractive and repulsive color channels , that come with an opposite sign in the grand potential .
it is tempting to conclude from this result that it provides an _ a posteriori _ justification of the success of approaches involving free quasiparticles in the description of the equation of state , even in the strongly coupled phase .
an obvious drawback of our framework is the neglect of chiral symmetry , leading to results that may be inaccurate in the light quark sector .
qcd in coulomb gauge is currently the formalism which is maximally close to ours while fully including chiral symmetry .
some work has been done in the study of pure yang - mills theory and by using a toy model with confining potential that mimics qcd @xcite .
modelling the full quark - gluon plasma within coulomb gauge qcd is however a huge task that still remains to be achieved .
some comments can be made about the large-@xmath148 behavior of our results .
the meson masses depend on the number of color through the factor @xmath236 only , so the meson masses are of order 1 at large @xmath148 , with corrections in @xmath237 as expected from a quenched potential .
moreover , it has been shown in @xcite that the quark contribution to the equation of state behaves as @xmath238 in t hooft s limit , as expected .
this is an important check of the ability of the present model to deal with the large-@xmath148 limit .
future developments of the present model should include the computation of the viscosity - over - entropy ratio .
such a computation can in principle be done without extra parameter .
hence , it is an important extension of our formalism that we hope to present in forthcoming works .
m. i. gorenstein and s. n. yang , phys .
d * 52 * , 5206 ( 1995 ) ; a. peshier , b. kmpfer , o. p. pavlenko , and g. soff , phys . rev .
d * 54 * , 2399 ( 1996 ) ; p. lvai and u. heinz , phys .
c * 57 * , 1879 ( 1998 ) . | the strongly - coupled phase of the quark - gluon plasma ( qgp ) is studied here by resorting to a @xmath0-matrix formulation in which the medium is seen as a non - ideal gas of quasiparticles ( quarks , antiquarks and gluons ) interacting nonpertubatively . in the temperature range under study , ( 1 - 5 ) @xmath1 , where @xmath1 is the temperature of deconfinement , the interactions are expected to be strong enough to generate bound states .
the dissociation temperature of such binary bound states is thus computed here .
the more the quasiparticles involved in the binary system are heavy , the more the bound state is likely to survive significantly above @xmath1 .
then , the qgp equations of state at zero and small baryonic potential are computed for @xmath2 and @xmath3 by resorting to the dashen , ma and bernstein formulation of statistical mechanics .
comparisons with current lattice qcd data are presented . |
late time accelerated expansion of the universe was indicated by measurements of distant type ia supernovae ( sne ia ) @xcite . this was confirmed by observations of cosmic microwave background ( cmb ) anisotropies by the wilkinson microwave anisotropy probe ( wmap ) @xcite , and the large - scale structure in the distribution of galaxies observed in the sloan digital sky survey ( sdss ) @xcite .
it is not possible to account for this phenomenon within the framework of general relativity containing only matter .
therefore , a number of models containing `` dark energy '' have been proposed as the mechanism for the acceleration .
there are currently many dark energy models , including cosmological constant , scalar field , quintessence , and phantom models @xcite .
however , dark energy , the nature of which remains unknown , has not been detected yet .
the cosmological constant , which is the standard candidate for dark energy , can not be explained by current particle physics due to its very small value , and it is plagued with fine - tuning problems and the coincidence problem .
an alternative method for explaining the current accelerated expansion of the universe is to extend general relativity to more general theories on cosmological scales . instead of adding an exotic component such as a cosmological constant to the right - hand side ( i.e. , the energy - momentum tensor ) of einstein s field equation , the left - hand side ( i.e. , the einstein tensor , which is represented by pure geometry ) can be modified .
typical models based on this modified gravity approach are @xmath1 models @xcite and the dvali gabadadze porrati ( dgp ) model @xcite ( for reviews , see @xcite ) . in @xmath1 models , the scalar curvature @xmath2 in the standard einstein
hilbert gravitational lagrangian is replaced by a general function @xmath1 .
by adopting appropriate function phenomenologically , @xmath1 models can account for late - time acceleration without postulating dark energy .
the dgp model is an extra dimension scenario . in this model ,
the universe is considered to be a brane ; i.e. , a four - dimensional ( 4d ) hypersurface , embedded in a five - dimensional ( 5d ) minkowski bulk . on large scales , the late - time acceleration is driven by leakage of gravity from the 4d brane into 5d spacetime .
naturally , there is no need to introduce dark energy . on small scales ,
gravity is bound to the 4d brane and general relativity is recovered to a good approximation . according to various recent observational data including that of type ia supernovae @xcite
, it is possible that the effective equation of state parameter @xmath3 , which is the ratio of the effective pressure @xmath4 to the effective energy density @xmath5 , evolves from being larger than @xmath6 ( non - phantom phase ) to being less than @xmath6 ( phantom phase @xcite ) ; namely , it has currently crossed @xmath6 ( the phantom divide ) .
@xmath1 models that realize the crossing of the phantom divide have been studied @xcite . on the other hand , in the original dgp model @xcite and a phenomenological extension of the dgp model described by the modified friedmann equation proposed by dvali and turner @xcite , the effective equation of state parameter never crosses the @xmath3 = @xmath6 line . in this paper
, we develop the `` phantom crossing dgp model '' by further extending the modified friedmann equation by dvali and turner @xcite . in our model ,
the effective equation of state parameter of dgp gravity crosses the phantom divide line , as indicated by recent observations .
this paper is organized as follows . in the next section ,
we summarize the original dgp model , and check the behavior of the effective equation of state . in section [ sec:3 ] , we describe the modified friedmann equation by dvali and turner @xcite , and we also demonstrate that the effective equation of state does not cross the @xmath3 = @xmath6 line in this framework . in section [ sec:4 ] , we construct `` the phantom crossing dgp model '' by extending the modified friedmann equation proposed by dvali and turner .
we show that the effective equation of state parameter of our model crosses the phantom divide line , and investigate the properties of our model .
finally , a summary is given in section [ sec:5 ] .
the dgp model @xcite assumes that we live on a 4d brane embedded in a 5d minkowski bulk .
matter is trapped on the 4d brane and only gravity experiences the 5d bulk .
the action is @xmath7 where the subscripts ( 4 ) and ( 5 ) denote quantities on the brane and in the bulk , respectively .
@xmath8 ( @xmath9 ) is the 5d ( 4d ) planck mass , and @xmath10 represents the matter lagrangian confined on the brane .
the transition from 4d gravity to 5d gravity is governed by a crossover scale @xmath11 .
@xmath12 on scales larger than @xmath11 , gravity appears 5d . on scales smaller than @xmath11
, gravity is effectively bound to the brane and 4d newtonian dynamics is recovered to a good approximation .
@xmath11 is the single parameter in this model . assuming spatial homogeneity and isotropy , a friedmann - like equation on the brane is obtained as @xcite @xmath13 where @xmath14 is the total cosmic fluid energy density on the brane .
@xmath15 represents the two branches of the dgp model .
the solution with @xmath16 is known as the self - accelerating branch . in this branch ,
the expansion of the universe accelerates even without dark energy because the hubble parameter approaches a constant , @xmath17 , at late times . on the other hand
, @xmath18 corresponds to the normal branch
. this branch can not undergo acceleration without an additional dark energy component .
hence in what follows we consider the self - accelerating branch ( @xmath16 ) only . for the second term on the right - hand side of eq .
( [ dgp_fri ] ) , which represents the effect of dgp gravity , the effective energy density is @xmath19 and the effective pressure is @xmath20 where @xmath21 , the differential of the hubble parameter with respect to the cosmological time @xmath22 . using eqs .
( [ rho_eff ] ) and ( [ p_eff ] ) , the effective equation of state parameter of dgp gravity is given by @xmath23 vs. redshift @xmath24 .
the red ( solid ) , green ( dashed ) , blue ( dotted ) lines represent the cases for @xmath25 , and @xmath26 , respectively ( corresponding to @xmath27 , and @xmath28 , respectively ) .
[ fig : dgp_wrc ] ] fig .
[ fig : dgp_wrc ] shows the behavior of the effective equation of state of dgp gravity @xmath29 versus the redshift @xmath24 for @xmath25 , and @xmath26 . assuming that the total cosmic fluid energy density @xmath14 of eq .
( [ dgp_fri ] ) contains matter and radiation , from eq .
( [ rch0omegam ] ) , these values of @xmath30 correspond to @xmath27 , and @xmath28 , respectively .
@xmath31 where @xmath32 is the normalized energy density of matter and @xmath33 is the radiation on the brane ; i.e. , @xmath34 and @xmath35 .
( @xmath36 , @xmath37 ) .
the subscripts @xmath38 designate the present value .
the effective equation of state of dgp @xmath29 can also be exactly expressed in terms of the energy densities of matter and radiation , @xmath32 and @xmath33 @xcite .
@xmath39 in realistic ranges of the energy density , @xmath40 and @xmath41 , the value of the effective equation of state can not be less than or equal to @xmath6 .
that is , the effective equation of state never crosses the phantom divide line in the original dgp model .
dvali and turner @xcite phenomenologically extended the friedmann - like equation ( eq .
( [ dgp_fri ] ) ) of the dgp model .
this model interpolates between the original dgp model and the pure @xmath42cdm model with an additional parameter @xmath43 .
the modified friedmann - like equation is @xcite @xmath44 for @xmath45 , this agrees with the original dgp friedmann - like equation , while @xmath46 leads to an expansion history identical to that of @xmath42cdm cosmology .
differentiating both sides of eq .
( [ dt_fri ] ) with respect to the cosmological time @xmath22 , we obtain the following differential equation .
@xmath47 where a dot indicates the derivative respect to the cosmological time .
the quantity @xmath48 is the total cosmic fluid pressure on the brane . for the second term on the right - hand side of eq .
( [ dt_fri ] ) , which represents the effect of dgp gravity , the effective energy density is @xmath49 and from eq .
( [ hdot2 ] ) the effective pressure is @xmath50 .
\label{p_dt}\ ] ] from eqs .
( [ rho_dt ] ) and ( [ p_dt ] ) , the effective equation of state parameter of the dgp model extended by dvali and turner is given by @xmath51 fig .
[ fig : dt_wa ] shows a plot of the behavior of the effective equation of state of the dgp model by dvali and turner @xmath52 versus the redshift @xmath24 for @xmath53 , and @xmath54 ( assuming @xmath55 ) .
, vs. redshift @xmath24 of the dgp model extended by dvali and turner for @xmath53 , and @xmath54 ( top to bottom ) assuming @xmath55 .
[ fig : dt_wa ] ] in general , for equation of state @xmath56 , the energy density @xmath14 varies as @xmath57 .
this leads to the following proportional relation .
@xmath58 at the same time , from eq .
( [ rho_dt ] ) , we find @xmath59 . in the radiation - dominated epoch , from the proportional relation on hubble parameter @xmath60
, we obtain the following relation . @xmath61 as compared the right - hand side of eq .
( [ rad1 ] ) to that of eq .
( [ rad2 ] ) , during the earlier radiation - dominated epoch ( @xmath62 ) , the effective equation of state can also be represented with @xmath43 @xcite .
@xmath63 in the same way , during the matter - dominated epoch ( @xmath64@xmath24@xmath65@xmath66 ) , from the proportional relation on hubble parameter @xmath67 , @xmath68 at the present time , @xmath52 is a stationary value close to @xmath6 . from these results , in the case of @xmath69 , the effective equation of state @xmath70 in all era , even in the radiation - dominated epoch .
that is to say , the component of the dgp gravity works as the driving force of the accelerated expansion of the universe in all epochs .
on the other hand , for @xmath71 , there is era when the effective equation of state becomes @xmath72 .
that is , the dgp gravity does not drive the accelerated expansion in all epochs .
the case of @xmath73 corresponds to the original dgp model described in the previous section .
thus , in the original dgp model , the effective equation of state @xmath74 in the radiation - dominated epoch . and
after the radiation - dominated epoch , becomes @xmath75 .
in other words , the dgp gravity acts as the driving force of the accelerated expansion just after the radiation - dominated epoch .
however , when @xmath43 is positive , the effective equation of state @xmath52 will exceed @xmath6 at all times . for negative @xmath43 , @xmath52 is always less than @xmath6 . in the case of @xmath46
, @xmath52 is @xmath6 constantly .
based on this analysis , crossing of the phantom divide does not occur in the dgp model extended by dvali and turner .
we propose the `` phantom crossing dgp model '' that extends the modified friedmann equation ( eq .
( [ dt_fri ] ) ) proposed by dvali and turner .
our model can realize crossing of the phantom divide line for the effective equation of state of the dgp gravity . as mentioned in the previous section , the effective equation of state parameter of the dgp model by dvali and turner @xmath52 ,
takes the value of over @xmath6 for positive @xmath43 , and it is less than @xmath6 for negative @xmath43 . when @xmath46 , @xmath52 becomes @xmath6 . on the basis of these results , we consider a model in which @xmath43 varies being positive to being negative . to keep the model as simple as possible , we make the following assumption , @xmath76 where @xmath77 is the scale factor ( normalized such that the present day value is unity ) .
the quantity @xmath0 is a constant parameter . in the period
when the scale factor @xmath77 is less than the parameter @xmath0 ( @xmath78 ) , the effective equation of state exceeds @xmath6 .
at the point when the scale factor @xmath77 equals @xmath0 , ( @xmath46 ) , the equation of state s value will be @xmath6 . in the period when the scale factor @xmath77 exceeds the parameter @xmath0 ( @xmath79 ) ,
the equation of state will be less than @xmath6 . in this way , crossing of the phantom divide is realized in our model .
replacing @xmath43 by @xmath80 in eq .
( [ dt_fri ] ) , the friedmann - like equation in our model is given by @xmath81 differentiating both sides of eq .
( [ hirano_fri ] ) with respect to the cosmological time @xmath22 , the following differential equation is obtained .
@xmath82 for the second term on the right - hand side of eq .
( [ hirano_fri ] ) representing the effect of dgp gravity , the effective energy density is @xmath83 and from eq .
( [ hirano_hdot2 ] ) , the effective pressure is @xmath84 .
\label{p_hirano}\ ] ] using eqs .
( [ rho_hirano ] ) and ( [ p_hirano ] ) , the effective equation of state of our model is given by @xmath85 , vs. redshift @xmath24 .
the red ( solid ) , green ( dashed ) , blue ( dotted ) lines represent the cases of @xmath86 , and @xmath87 , respectively ( assuming @xmath55 ) .
[ fig : dgp_hirano],width=445 ] depicted in fig .
[ fig : dgp_hirano ] near recent epochs .
[ fig : dgp_hirano_kakudai],width=445 ] fig .
[ fig : dgp_hirano ] shows a plot of the effective equation of state of our model @xmath88 versus the redshift @xmath24 ( see also fig .
[ fig : dgp_hirano_kakudai ] which shows an enlarged view of this diagram ) .
our model is an extension of the dgp model and realizes crossing of the phantom divide .
the effective equation of state @xmath88 of models for @xmath89 and @xmath87 ( assuming @xmath55 ) crosses the phantom divide line when the redshift @xmath90 and @xmath91 , respectively .
we find that the smaller the parameter @xmath0 is , the older epoch crossing of the phantom divide occurs in .
@xmath0 is not necessarily equal to the scale factor at the time of crossing the phantom divide , even though eq .
( [ beta ] ) is assumed . in the eq .
( [ hirano_fri ] ) , the value of @xmath80 that is the power index of @xmath92 varies with respect to time . as the power index of the differential equation changes with time , furthermore , in parallel , the differential equation is solved with respect to time .
hence , the time lag occurs , the scale factor at the time of crossing the phantom divide is more than the value of @xmath0 . in a way similar to the derivation of eq .
( [ wa_rad ] ) , we represent the effective equation of state of phantom crossing dgp model with @xmath0 . in the radiation - dominated epoch , the scale factor @xmath77 is taken to be @xmath38 in comparison with the value of @xmath0 .
therefore , as @xmath93 in eq .
( [ beta ] ) , during the radiation - dominated epoch ( @xmath62 ) , the effective equation of state is approximately @xmath94 that is , in the case of @xmath95 , the effective equation of state @xmath96 in all era , including the radiation - dominated epoch . on the other hand , for @xmath97
, there is era when the effective equation of state becomes @xmath98 . and absolute value of the effective pressure @xmath99 ( note that @xmath100 ) of our model vs. redshift @xmath24 , for ( @xmath101 ) = ( @xmath102 ) .
[ fig : rhop_hirano ] ] fig .
[ fig : rhop_hirano ] shows the effective energy density @xmath103 and absolute value of the effective pressure @xmath99 ( note that @xmath100 ) of our model for ( @xmath101 ) = ( @xmath102 ) versus the redshift @xmath24 , normalized such that the effective energy density is unity at the time of phantom crossing .
it shows that the absolute value of the effective pressure @xmath99 exceeds the effective energy density @xmath103 at the time of crossing of the phantom divide .
the recent observational data for type ia supernovae @xcite show that crossing of the phantom divide line occurs at a redshift @xmath104 @xcite . in our model , for @xmath105 ( when @xmath55 ) , crossing of the phantom divide occurs at @xmath104 .
+ in a proposed model in which the phantom divide is crossed at @xmath104 , ( @xmath101 ) = ( @xmath102 ) , we investigate and show the property of phantom crossing dgp model .
relative to that of a constant expansion cosmology @xmath106 , vs. the redshift @xmath24 . models and parameters
are ( from top to bottom ) : ( 1 ) @xmath42cdm model , @xmath107 = 0.30 ; ( 2 ) phantom crossing dgp model , @xmath0 = 0.50 , @xmath107 = 0.30 ; ( 3 ) dgp model by dvali and turner , @xmath43 = 0.50 , @xmath107 = 0.30 ; ( 4 ) original dgp model , @xmath107 = 0.30 .
[ fig : snia ] ] , matter @xmath32 , and dgp gravity @xmath108 , vs. the redshift @xmath24 in the phantom crossing dgp model with the proposed parameter ( @xmath101 ) = ( @xmath102 ) .
[ fig : omega ] ] fig .
[ fig : snia ] shows the distance modulus @xmath109 relative to that of a constant expansion cosmology @xmath106 , versus the redshift @xmath24 .
that is , when @xmath110 is positive , cosmic expansion is accelerating .
the distance modulus is defined by @xmath111 where @xmath112 is the hubble free luminosity distance given by @xmath113 @xmath114 being the hubble constant @xmath115 in units of @xmath116 .
we adopt @xmath117 @xcite . in fig .
[ fig : snia ] , models and parameters are ( from top to bottom ) : ( 1 ) @xmath42cdm model , @xmath107 = 0.30 ; ( 2 ) phantom crossing dgp model , @xmath0 = 0.50 , @xmath107 = 0.30 ; ( 3 ) dgp model by dvali and turner , @xmath43 = 0.50 , @xmath107 = 0.30 ; ( 4 ) original dgp model , @xmath107 = 0.30
. phantom crossing dgp model can realize late - time acceleration of the universe very similar to that for @xmath42cdm model , without dark energy .
[ fig : omega ] shows the normalized energy density of radiation @xmath33 , matter @xmath32 , and dgp gravity @xmath108 versus the redshift @xmath24 in the phantom crossing dgp model with the proposed parameter ( @xmath101 ) = ( @xmath102 ) .
where , @xmath118 .
@xmath119 is the effective energy density of dgp gravity defined by eq .
( [ rho_hirano ] ) .
we find that the universe is dgp gravity - dominated near recent epochs .
therefore , in the phantom crossing dgp model , the late - time acceleration is driven by the effect of dgp gravity .
, vs. the redshift @xmath24 .
models and parameters are same as fig . [
fig : snia ] .
[ fig : weff ] ] fig .
[ fig : weff ] shows the effective equation of state @xmath120 versus the redshift @xmath24 .
models and parameters are same as fig .
[ fig : snia ] .
only our phantom crossing dgp model can realize crossing of the phantom divide line at @xmath104 as indicated by recent observations .
* we confirmed that the effective equation of state does not cross the phantom divide line in the original dgp model . + * we also demonstrated that crossing of the phantom divide does not occur in the dgp model by dvali and turner . +
* we constructed the phantom crossing dgp model .
this model realizes crossing of the phantom divide .
we found that the smaller the value of the new introduced parameter @xmath0 is , the older epoch crossing of the phantom divide occurs in .
our model can realize late - time acceleration of the universe very similar to that of @xmath42cdm model , without dark energy , due to the effect of dgp gravity . in the proposed model ( ( @xmath101 ) = ( @xmath102 ) ) , crossing of the phantom divide occurs at @xmath104 as indicated by recent observations .
riess , et al .
j. * 116 * , 1009 ( 1998 ) s. perlmutter , et al .
, astrophys .
j. * 517 * , 565 ( 1999 ) r. knop , et al .
, astrophys .
j. * 598 * , 102 ( 2003 ) a.g .
riess , et al .
, astrophys .
j. * 607 * , 665 ( 2004 ) a.g .
riess , et al .
, astrophys .
j. * 659 * , 98 ( 2007 ) p. astier , et al .
astrophys . * 447 * , 31 ( 2006 ) g. miknaitis , et al .
, astrophys .
j. * 666 * , 674 ( 2007 ) w.m .
wood - vasey , astrophys .
j. * 666 * , 694 ( 2007 ) j.a .
frieman , et al .
j. * 135 * 338 ( 2008 ) d.n .
spergel , et al .
, astrophys .
* 148 * , 175 ( 2003 ) e. komatsu , et al .
[ wmap collaboration ] , astrophys .
* 180 * , 330 ( 2009 ) m. tegmark , et al .
[ sdss collaboration ] , astrophys .
j. * 606 * , 702 ( 2004 ) m. tegmark , et al .
d * 69 * , 103501 ( 2004 ) b. ratra , p.j.e .
peebles , phys .
d * 37 * , 3406 ( 1988 ) p.j.e .
peebles , b. ratra , astrophys .
j. * 325 * , l17 ( 1988 ) r.r .
caldwell , phys .
b * 545 * , 23 ( 2002 ) r.r .
caldwell , m. kamionkowski , n.n .
weinberg , phys .
lett * 91 * , 071301 ( 2003 ) k. hirano , k. kawabata , z. komiya , astrophys .
space sci . * 315 * , 53 ( 2008 ) j.g .
hartnett , k. hirano , astrophys .
space sci . * 318 * , 13 ( 2008 ) k. hirano , et al . ,
proceedings of 59th yamada conference `` inflating horizons of particle astrophysics and cosmology '' ( universal academy press ) , p.219 ( 2005 ) z. komiya , k. kawabata , k. hirano , h. bunya , n. yamamoto , j. korean astron .
soc . * 38 * , 157 ( 2005 ) z. komiya , k. kawabata , k. hirano , h. bunya , n. yamamoto , astron .
astrophys . * 449 * , 903 ( 2006 ) s. nojiri , s.d .
odintsov , int .
phys . * 4 * , 115 ( 2007 ) s. nojiri , s.d .
odintsov , arxiv:0801.4843 [ astro - ph ] .
s. nojiri , s.d .
odintsov , arxiv:0807.0685 [ hep - th ] .
dvali , g. gabadadze , m. porrati , phys .
b * 485 * , 208 ( 2000 ) c. deffayet , phys .
b * 502 * , 199 ( 2001 ) c. deffayet , g.r .
dvali , g. gabadadze , phys .
d * 65 * , 044023 ( 2002 ) k. koyama , gen .
. grav . * 40 * , 421 ( 2008 ) .
u. alam , v. sahni , a.a .
starobinsky , jcap * 0406 * , 008 ( 2004 ) s. nesseris , l. perivolaropoulos , jcap * 0701 * , 018 ( 2007 ) p.u .
yu , phys .
b * 643 * , 315 ( 2006 ) h.k .
jassal , j.s .
bagla , t. padmanabhan , astro - ph/0601389 .
s. nojiri , s.d .
odintsov , phys .
b * 562 * , 147 ( 2003 ) k. bamba , c.q .
geng , phys .
b * 679 * , 282 ( 2009 ) k. bamba , c.q .
geng , s. nojiri , s.d .
odintsov , phys .
d * 79 * , 083014 ( 2009 ) g. dvali , m. turner , astro - ph/0301510 .
k. nozari , m. pourghassemi , jcap * 0810 * , 044 ( 2008 ) k. nozari , n. behrouz , t. azizi , b. fazlpour , prog .
. phys . * 122 * , 735 ( 2009 ) a. lue , r. scoccimarro , g.d .
starkman , phys .
d * 69 * , 124015 ( 2004 ) a. lue , phys
. rep . * 423 * , 1 ( 2006 ) w.l .
freedman , et al .
, astrophys .
j. * 553 * , 47 ( 2001 ) | we propose a phantom crossing dvali gabadadze porrati ( dgp ) model . in our model ,
the effective equation of state of the dgp gravity crosses the phantom divide line .
we demonstrate crossing of the phantom divide does not occur within the framework of the original dgp model or the dgp model developed by dvali and turner . by extending their model , we construct a model that realizes crossing of the phantom divide .
we find that the smaller the value of the new introduced parameter @xmath0 is , the older epoch crossing of the phantom divide occurs in .
our model can account for late - time acceleration of the universe without dark energy .
we investigate and show the property of phantom crossing dgp model .
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 |
black holes span a wide spectrum of masses : the case for stellar mass holes ( @xmath0 ) in the field [ e.g. , @xcite ] and supermassive holes ( @xmath1 ) in galactic bulges [ e.g. , @xcite ] is extremely strong ; tantalizing evidence suggests middleweight holes ( @xmath2 ) as well @xcite .
stellar mass holes likely form in stellar collapse ; the origins of more massive holes remains mysterious .
such holes could form in the collapse of massive gas accumulations ; they could grow from smaller holes by accretion ; they could grow by capturing stellar mass bodies ; and they could grow by repeatedly merging with holes of comparable mass .
any or indeed all of these mechanisms could contribute to the growth of a given hole .
a black hole s spin may help identify which scenario most strongly impacted its recent history . since spin likely drives outflows and jets in active galaxies , and since jets are presumed to align with black hole spin @xcite , spin may provide an observational probe of a hole s recent growth @xcite .
we examine how spin and mass coevolve in mergers .
binaries will form following galaxy mergers @xcite , and may harden to the point that gravitational - wave ( gw ) emission drives its members together .
eventually , they encounter the last stable orbit ( lso ) , and then plunge and coalesce into a single hole .
our goal is to understand the mass and spin of this remnant hole . for
nearly equal mass holes , this is extremely difficult : we must model the spacetime dynamics of the transition from a binary to a single black hole , accounting for both holes spins and the radiated energy and angular momentum .
a proper analysis requires mature numerical relativity codes [ see , e.g. , @xcite ] .
the problem is simpler for small mass ratio , @xmath3 .
this binary is well described as a test particle orbiting a black hole .
gw emission shrinks the small hole s orbit to the lso , whereupon it plunges into the large hole .
neglecting the final emission of radiation after the plunge , the hole evolves simply : its mass adds the small body s energy at the lso , its spin adds the lso angular momentum . because we only need
global `` conserved '' quantities , this description works surprisingly well even for rather large mass ratio .
post - newtonian analyses @xcite show that finite mass ratio typically changes the lso and its orbital constants by a factor of order @xmath4 .
the error due to the test particle description is @xmath5 for @xmath6 .
we also may safely neglect the energy and angular momentum radiated in the final merger : although its gw luminosity may be large , its duration will be very short .
the mass carried off in this phase , for example , is @xmath7 @xcite . neglecting this radiation incurs an error that is less important than other errors built into our approximations , and rapidly becomes negligible for small mass ratio .
likewise , the small hole s spin can be neglected : since a hole s spin scales with its mass squared , spin will be less important than the orbital angular momentum , provided we exclude @xmath8 .
we set the speed of light @xmath9 and newton s constant @xmath10 to 1 ; a useful conversion factor is @xmath11 .
our binary has masses @xmath12 and
@xmath13 ; the larger hole has mass @xmath12 and spin @xmath14 ; @xmath15 .
vectors are written in boldface ; hatted quantities have been made dimensionless by dividing out powers of mass e.g. , @xmath16 .
on reaching the lso , the smaller hole plunges into the large hole , carrying along its orbital constants energy @xmath17 , angular momentum parallel to the spin @xmath18 , and `` carter constant '' @xmath19 .
the carter constant separates the equations of motion in a hamilton - jacobi description of black hole orbits [ e.g. , @xcite ( mtw ) ] .
it is essentially just the `` rest '' of the orbit s angular momentum : to very high accuracy @xcite , one can describe the binary as having an angular momentum @xmath20 .
we treat @xmath19 as @xmath21 , angular momentum projected into the equatorial plane ( perpendicular to the spin ) .
this treatment resonates with the theory of orbits in axisymmetric potentials : @xmath19 is a relativistic analog of the `` 3rd integral '' @xmath22 [ cf .
@xcite ] . treating @xmath19 as @xmath21 is exact for orbits of non - spinning holes ; for maximal spin , the error is less than a few percent @xcite .
@xmath17 includes the rest mass of the orbiting body : bound orbits have @xmath23 , unbound orbits @xmath24 .
the lso is a one - parameter set of orbits : each orbit has radius @xmath25 and constants @xmath26 determined by the inclination angle @xmath27 , defined as @xmath28 this angle is very useful : detailed studies @xcite show it remains practically constant during inspiral , so a distribution @xmath29 describing an ensemble of binaries at formation likewise describes that ensemble at plunge . to find the constants for circular orbits at plunge , solve @xmath30
prime denotes @xmath31 , and the `` potential '' @xmath32 is given by @xmath33 ^ 2- \delta(r)\left[r^2 + ( l_z - ae)^2 + q\right]\ ; , \label{eq : radpotential}\ ] ] where @xmath34 .
this potential describes the orbit s radial motion ; see mtw , chap .
the lso is bounded by prograde ( @xmath35 ) and retrograde ( @xmath36 ) equatorial orbits , with constants @xcite @xmath37 where @xmath38 , and @xmath39\;,\\ z_2 & = & \left[3 { \hat a}^2 + z_1 ^ 2\right]^{1/2}\;.\end{aligned}\ ] ] the upper sign is for prograde orbits , the lower for retrograde .
we now drop the `` lso '' subscript , since we always refer to these quantities at the lso . using the bounding cases as initial guesses , it is straightforward to solve eq .
( [ eq : reqns_circ ] ) using newton s method @xcite for the constants .
the results are surprisingly well fit by a simple rule : letting @xmath40 stand for @xmath25 , @xmath17 , or @xmath41 , @xmath42 using eq .
( [ eq : xi_lso ] ) for @xmath41 , one builds @xmath43 , @xmath44 .
these fits are extremely good for small spin , and induce errors of about @xmath45 for @xmath46 .
we should also consider eccentricity @xmath47 .
major mergers are well described by a circular lso ( eccentricity rapidly bleeds away in the formation of a tight binary and by the inspiral ) , but minor mergers will have significant eccentricity .
we find that a black hole growth history barely changes when eccentricity is taken into account : almost identical growth histories are obtained with @xmath48 and with @xmath49 .
we will confine our discussion to circular orbits .
it is now simple to compute a remnant s properties . before merger , the large hole has mass @xmath12 and spin @xmath50 along the @xmath51 axis .
after merger , the remnant has mass and spin @xmath52\ ; , \label{eq : mnew}\\ s'_z & = & m^2[{\hat a } + q { \hat l}_z({\hat a},\mu)]\ ; , \label{eq : sznew}\\ s'_\perp & = & q m^2 { \hat l}_\perp({\hat a},\mu)\;. \label{eq : sperpnew}\end{aligned}\ ] ] the remnant hole is inclined at an angle @xmath53 relative to the original hole , and has spin @xmath54 : @xmath55
we now examine the remnant s properties following a single merger , choosing @xmath56 and computing @xmath26 as functions of the larger hole s spin @xmath57 and the inclination cosine @xmath58 .
we do not yet use the approximation ( [ eq : xi_lso ] ) , but instead solve eq.([eq : reqns_circ ] ) numerically .
we then use eqs .
( [ eq : mnew ] ) ( [ eq : anew ] ) to describe the remnant .
two examples , @xmath59 and @xmath60 , are shown in fig .. we show @xmath54 as a function of @xmath61 and @xmath58 .
consider the @xmath59 results first .
for a broad range of @xmath57 and @xmath58 , the remnant spins relatively slowly .
in many cases , @xmath62 : over much of the parameter space , this hole is spun _ down _ by the merger .
rapid rotation follows only if the larger hole was already spinning rapidly and plunge occurred at shallow inclination ( @xmath63 ) . nearly nonspinning remnants form near @xmath64 , @xmath65 :
the retrograde orbit cancels the hole s spin .
this requires @xmath66 when @xmath67 , the spin orientation changes very little by the merger .
by contrast , when @xmath68 , the spin is overwhelmed by the plunging body , and the orientation aligns with the plunge angle .
whereas about half of the parameter space for @xmath59 leads to remnants with spin @xmath69 , the half - area contour when @xmath60 is at spin @xmath70 . the slow spin of most remnants spin is simply understood .
angular momentum at the lso depends strongly on inclination cf . the fit ( [ eq : xi_lso ] ) .
the magnitude @xmath71 is small for prograde orbits and large for retrograde orbits .
thus @xmath71 is smallest when it tends to augment the spin ( @xmath72 and @xmath73 nearly parallel ) and is largest when it tends to cancel ( @xmath72 and @xmath74 nearly antiparallel ) . on average , the hole tends to spin down .
this tendency breaks at large @xmath56 , when @xmath73 overwhelms @xmath75 .
the spin of the remnant is then dominated by the orbit at plunge .
after a hole grows above a certain mass , its subsequent spin evolution may be stochastic .
this limit may be described by a fokker - planck equation , combining the secular and diffusive changes in the hole s characteristics .
we treat @xmath76 and @xmath77 as independent variables , so that the distribution function ( per unit volume @xmath77 space ) @xmath78 evolves via @xmath79 where @xmath80 is the @xmath81-th component of @xmath77 , @xmath82 and angle brackets mean to average over its distribution [ see , e.g. , pathria ( 1972 ) , lifshitz & pitaevski ( 1980 ) ] .
we restrict our attention to a population of black holes that are born with a specific mass and spin .
this solution can be used to integrate over a broad initial population .
we also will limit our quantitative analysis to an isotropic distribution of ( small ) merging holes ; generalization to an anisotropic distribution is straightforward though lengthy .
we rewrite eq .
( [ eq : fp ] ) in spherical coordinates attached to the initial spin direction , @xmath83 . for the isotropic case
, symmetry dictates that the only non - zero coefficients are @xmath84 equation ( [ eq : fp ] ) then becomes @xmath85 \nonumber\\ & + & { 1\over2}{d_{\perp}\over{\hat a}^2\sin\theta } { \partial\over\partial\theta}\sin\theta{\partial f\over\partial\theta}\;. \label{eq : fpiso}\end{aligned}\ ] ] evaluating the coefficients requires that we expand eqs.([eq : mnew ] ) ( [ eq : anew ] ) to leading order in @xmath56 . for each merger ,
the changes in @xmath86 and @xmath87 satisfy @xmath88 + o(q^2)\ ; , \label{eq : delta_ahat}\\ \delta\theta & = & { q\over{\hat a}}{\hat l}({\hat a},\mu)\sqrt{1 - \mu^2 } + o(q^2)\;. \label{eq : delta_theta}\end{aligned}\ ] ] combining this with @xmath89 yields @xmath90\over { \hat e}(\hat a,\mu)}\right\rangle \approx -2.4 { \hat a}\ ; ; \label{eq : rterm}\\ d_{|| } \!\!\!\!&=&\!\!\!\ ! \langle q\rangle\left\langle{[{\hat l}_z({\hat a},\mu ) - 2{\hat a}]^2\over { \hat e}(\hat a,\mu)}\right\rangle \approx \langle q\rangle(4 + 4.9 { \hat a}^2)\ ; ; \label{eq : dparterm}\\ d_{\perp } \!\!\!\!&=&\!\!\!\ ! { \langle q\rangle\over2}\left\langle{[{\hat l}({\hat a},\mu)\sqrt{1 - \mu^2}]^2\over{\hat e}(\hat a,\mu)}\right\rangle \approx \langle q\rangle(4 - 0.9{\hat a}^2 ) .
\label{eq : dperpterm}\end{aligned}\ ] ] @xmath91 is the typical minor merger mass ratio . in the approximations , we put @xmath92 and expand in powers of @xmath93 ; this is permissible for @xmath94 .
the fokker - planck approach works well for @xmath95 , and is easily supplemented by direct sums ( as in sec .
[ subsec : major ] ) for major mergers .
the fokker - planck coefficients can be used to understand semi - quantitatively the evolution of holes that grow through minor mergers .
the change in @xmath77 is dominated by the resistive term for @xmath96 , which is the case for @xmath97 . in this case , the fluctuations about the average evolution are small .
this average , secular evolution can be found by using the definition of @xmath32 and integrating eq .
( [ eq : rterm ] ) to find @xmath98^{2.4}\;. \label{eq : secularevolve}\ ] ] if the power were @xmath99 rather than @xmath100 , this equation would tell us that the hole s original spin is preserved while the mass grows .
the spin dies away somewhat faster because the magnitude of the change for retrograde captures is larger than that of prograde captures .
the orientation evolves diffusively .
after growing by @xmath101 , the typical misalignment is @xmath102 significant realignment in a single merger occurs only if @xmath103 , in accord with eq .
( [ eq : qcrit ] ) .
we have argued that a hole is unlikely to acquire rapid spin through mergers .
turn to the converse problem : if we believe that a black hole is spinning with @xmath104 , how many minor mergers can it have experienced ?
to be specific , suppose that the hole is born through gravitational collapse or is spun up by accretion to @xmath105 .
if we assume that the hole spun down to its current value by capture from an isotropic distribution of minor mergers , then as @xmath106 for @xmath107 , the mass acquired must satisfy @xmath108 rapidly spinning holes are extremely unlikely to have suffered a recent merger ; their spin must either be original or due to accretion .
we find that the remnant of a major merger is rarely rapidly rotating : rapid rotation follows only if the larger binary member spun rapidly before merger _ and _ the plunge was nearly prograde ; or , if the binary s mass ratio @xmath109 .
given the variety of black hole masses seen in galaxies , mergers with @xmath109 should be rare ; given the small volume of parameter space leading to rapid rotation , serendipitous configurations leading to a rapid rotation should also be rare .
mounting evidence , mostly from spiral galaxies , suggests that in many cases massive black holes nonetheless rapidly rotate [ e.g. , @xcite ] .
our results strongly suggest that this spin can not come from mergers [ e.g. @xcite ] but , instead , is consistent with the view that black hole mass ( and spin ) is assembled radiatively [ e.g.@xcite ] .
the spin evolution of a hole that grows by repeated minor mergers is neatly described by a fokker - planck equation [ eq . ( [ eq : fp ] ) ] , taking a particularly simple form if the mergers arise from an isotropic cluster . in this case
, the evolution has a secular component , which is approximately described by a `` doctrine of original spin '' [ @xmath110 remains roughly constant while @xmath12 grows ; cf .
( [ eq : secularevolve ] ) ] , and a diffusive component , with a spectrum of fluctuations governed by the coefficient @xmath111 [ cf .
( [ eq : dparterm ] ) ] .
this is in accord with recent work on models to grow intermediate mass black holes in clusters @xcite .
finally , we predict the typical angle @xmath112 by which a hole s orientation changes following merger [ eq.([eq : delta_theta ] ) ] .
these results may be of particular observational interest .
jets launched by a black hole s spin should track its inclination : if the hole is kicked into a new orientation , the jet will `` kink '' .
the angle of the kink should equal the hole s change in orientation . applying our predicted dependence for the kink angle on the binary s parameters may provide some insight into the conditions of the hole
s last merging , untangling a bit of its recent growth history . an abrupt change in inclination [ such as discussed in @xcite ] requires a comparatively rare major merger . this work grew out of discussions at the kitp conference _
black holes : theory confronts reality , three years later _ ; we thank the participants of that meeting for their stimulating input .
we are also extremely grateful to xinwu cao for pointing out that the figure used in an earlier version of this manuscript contained important errors .
sah is supported by nsf grant phy-9907949 ; rdb is supported by nasa grants nagw 5 - 2837 , 5 - 7007 .
van der marel , r. p. 2001 , in black holes in binaries and galactic nuclei .
proceedings of the eso workshop held at garching , germany , 6 - 8 september 1999 , eds .
l. kaper , e. p. j. van den heuvel , p. a. woudt , 246 ( astro - ph/0001342 ) | massive black holes appear to be present in the nuclei of almost all galaxies , but their genesis and evolution are not well understood . as astrophysical black holes
are completely characterized by their masses and spins , the observed joint distribution of these quantities contains important clues to their history .
we examine the coevolution of mass and spin in binary merger growth scenarios .
we find that holes are typically spun _ down _ by mergers .
rapid rotation results only if the binary s larger member already spins quickly and the merger with the smaller hole is consistently near prograde ; or , if the binary s mass ratio approaches unity .
if , as some observations have suggested , observed black holes spin rapidly , then this limits the importance of merger scenarios for the growth of black holes . |
( snr ) is an important index for wireless communication systems . in wireless communication systems , it is the most significant to achieve high capacity @xcite . in general , it is necessary and sufficient for achieving high capacity to increase snr under the condition that the width of the frequency band is constant @xcite .
similary , the performance of wireless communication is evaluated in bit error rate ( ber ) .
however , these two are not independent , and it is known that ber decreases as snr increases . as a wireless communication system ,
we focus on a code division multiple access ( cdma ) system @xcite , in particular , an asynchronous cdma system .
it is one of the multiple access systems with which many people can communicate each other at the same time @xcite . in cdma systems , spreading sequences are utilized as codes to multiplex . each user is assigned a different code and uses it to modulate and demodulate his signal . in cdma systems ,
many methods have been proposed to increase snr .
the one of such methods is based on the blind multiuser detection @xcite . on the other hand , improving the receiver @xcite with the application of digital implementation of ica @xcite and maximum likelihood ( ml ) estimation @xcite are also efficient .
however , in particularly , ml estimation method needs a large amount of calculations . on the contrary , to increase snr
, the representative approach is to improve spreading sequences .
the current spreading sequence of 3 g cdma systems is the gold code @xcite .
it is known that the gold code is optimal in all the binary spreading sequences as well as the kasami sequence @xcite . to explore a better sequence , in @xcite and @xcite ,
the use of chaotic spreading sequences has been proposed .
these chaotic spreading sequences are multivalued sequences , not binary ones , and are obtained from chaotic maps .
examples of such spreading sequences have been given in @xcite-@xcite .
however , the spreading sequences whose snr is maximum in all the spreading sequences are not yet obtained . in @xcite , the approach to obtain the capacity of spreading sequences has been proposed .
however , the sequence achieving maximal capacity has not been suggested . to achieve the maximal capacity
, we have to derive the practical spreading sequences whose interference noise is minimal . in general , crosscorrelation is treated as a basic component of interference noise , and autocorrelation is related to synchronization at the receiver side and the fading noise .
thus , it is desirable that the first peak of crosscorrelation and the second peak in autocorrelation should be kept low .
however , sarwate @xcite has shown that there is an avoidable limitation trade - off as a relation between lowering crosscorrelation s peak and autocorrelation s second peak . from this result , it is impossible that both of the peaks of crosscorrelation and autocorrelation are zero .
welch @xcite shows that the maximum value of crosscorrelation has a universal lower bound . from this result , it is impossible that the maximum value of crosscorrelation is zero in some situations . therefore , it is not straightforward to derive practical spreading sequences whose snr is high . in @xcite , sarwate has shown two kinds of characterized sequences on his limitation .
one kind is a set of sequences whose periodic crosscorrelation is always zero .
we call them sarwate sequences .
the other kind is a set of sequences whose periodic autocorrelation is always zero except for only one point , that is , frank - zadoff - chu ( fzc ) sequences @xcite @xcite . in @xcite , the extended set of the fzc sequences ,
the fzc sequence families are proposed .
they have three parameters and their snr , autocorrelation and crosscorrelation has been investigated . in this paper
, we define the weyl sequence class , which is a set of sequences generated by the weyl transformation @xcite .
this class belongs to the extended fzc sequence families and includes the sarwate sequences .
the sequence in the weyl sequence class has a desired property that the order of crosscorrelation is low .
we evaluate the upper bound of crosscorrelation and construct the optimization problem : minimize the upper bound of crosscorrelation . from the problem ,
we derive optimal spreading sequences in the weyl sequence class .
we show snr of them in a special case and compare them with other sequences in bit error rate .
in this section , we define the weyl sequence class and show their properties . let @xmath0 be the length of spreading sequences .
we define the weyl sequence @xmath1 as the following formula @xcite @xmath2 where @xmath3 and @xmath4 are real parameters . from the above definition
, we can assume that the parameters @xmath3 and @xmath4 satisfy @xmath5 and @xmath6 .
the sequences whose @xmath3 is an irrational number are used in a quasi - monte carlo method @xcite .
we apply this sequence to a spreading sequence .
then , the weyl spreading sequence @xmath7 is defined as @xcite @xmath8 where @xmath9 is the number of the user and @xmath10 is the unit imaginary number . in cdma systems ,
the value of @xmath11 has no effects to signal to noise ratio ( snr ) since @xmath12 is united to the phase term of the signal .
thus , we set @xmath13 .
we call the class which consists of weyl spreading sequences as the weyl sequence class .
note that this class is similar to the fzc sequence families @xcite .
the @xmath14-th element of the fzc sequence families is defined as @xmath15 where @xmath16 is an integer that is relatively prime to @xmath0 such that @xmath17 and @xmath18 and @xmath19 are any real numbers .
the triple @xmath20 specifies the set of sequences .
when the triple @xmath20 is @xmath21 , we obtain the element of the fzc sequence @xcite@xcite @xmath22 the weyl sequence class is obtained when the triple is @xmath23 and @xmath24 .
note that @xmath16 is not always an integer .
thus , the weyl sequence class belongs to the extended fzc sequence families whose @xmath16 is a real number . the element of the weyl sequence class , @xmath7 has a desired property that crosscorrelation is low .
we define the periodic correlation function @xmath25 and aperiodic correlation function @xmath26 as @xmath27 where @xmath28 and @xmath29 is the conjugate of @xmath30 .
the correlation functions @xmath25 and @xmath26 have been studied in @xcite @xcite @xcite . when @xmath31 , @xmath25 and @xmath26 are periodic and aperiodic crosscorrelation functions .
it is necessary for achieving high snr to keep the value of the crosscorrelation functions , @xmath32 and @xmath33 low for all @xmath34 . the absolute value of crosscorrelation functions @xmath32 and @xmath33 have the common upper bound that @xmath35 as @xmath36 the equality is attained if and only if @xmath37 where @xmath38 . from the above result
, @xmath39 obeys @xmath40 similarly , @xmath41 obeys @xmath42 .
thus , the upper bound of @xmath32 and @xmath33 is independent of @xmath0 . for
the general spreading sequences , due to the central limit theorem ( clt ) , the crosscorrelations @xmath32 and @xmath33 become large as @xmath0 becomes large .
for this reason , compared to the general spreading sequences , the weyl spreading sequence is expected to have low crosscorrelation .
in this section , we consider an asynchronous binary phase shift keying ( bpsk ) cdma system . our goal is to derive the spreading sequences whose interference noise is the smallest in the weyl sequence class .
let @xmath43 , @xmath44 and @xmath45 be the number of users , the durations of the symbol and each chip , respectively . in this situation , the user @xmath46 despreads the spreading sequences @xmath7 with the spreading sequence of the user @xmath46 , @xmath47 .
the symbols @xmath48 denote bits which the user @xmath9 send .
the transmitted signal of the user @xmath9 has time delay @xmath49 . from @xcite
, we assume that time delay @xmath49 is distributed in @xmath50 and satisfies @xmath51 , where @xmath52 is an integer .
then , the interference noise between the user @xmath46 and the user @xmath9 , @xmath53 is obtained as @xmath54 , \label{eq : correlation } \end{split}\ ] ] where @xmath55 is the phase of user @xmath9 s carrier . with eq .
( [ eq : bound1 ] ) , the absolute value of the interference noise @xmath53 is evaluated as @xmath56 thus , we have shown that the upper bound of interference noise between two sequences is inversely proportional to @xmath57 . to reduce the interference noise @xmath53 ,
it is necessary to reduce @xmath58 . to eliminate the absolute value function
, we introduce the distance between the phases @xmath59 and @xmath60 .
the distance @xmath61 we propose here is given by @xmath62 note that this @xmath63 satisfies the axiom of distance , and @xmath64 @xmath65 if we regard @xmath66 in the same light as @xmath67 . from eq .
( [ eq : d ] ) , we rewrite eq .
( [ eq : bound2 ] ) without any absolute value as @xmath68 we should take into account the whole interference noise in the users .
the whole interference noise @xmath69 is written as @xmath70 with eq .
( [ eq : bound3 ] ) , it is clear that @xmath71 has the upper bound : @xmath72 thus , we minimize eq .
( [ eq : whole ] ) and obtain the problem @xmath73 @xmath74 this problem is equivalent to that we minimize the sum of the upper bound of @xmath75 .
thus , the crosscorrelation among all the users is expected to be always low when we solve this problem . from eq .
( [ eq : d_prop1 ] ) , it is clear that @xmath76 .
then , in the problem @xmath73 , we count two times the same distance .
thus , we obtain the equivalent problem @xmath77 @xmath78 it is not clear if the objective function of the problem @xmath77 is convex since the form of function @xmath63 is complicated . to eliminate the function @xmath63 , we introduce slack variables @xmath79 for @xmath80 .
then , the problem @xmath77 is rewritten as @xmath81 without loss of generality , we assume @xmath82 . then , the problem @xmath83 can be rewritten as @xmath84 notice that the objective function and the inequality constraints of the problem @xmath80 are convex . it has been known that convex programming can be solved with the kkt conditions @xcite . to write such conditions ,
we define the variable vector @xmath85 as @xmath86 where @xmath87 , @xmath88 , @xmath89 and @xmath90 is the transpose of @xmath91 . from the kkt conditions ,
the solution @xmath92 is a global solution of @xmath80 if @xmath92 satisfies the following equation : @xmath93 where @xmath94 and the lagrange multipliers @xmath95 and @xmath96 are non - negative real numbers .
they have to satisfy the following conditions : @xmath97 in the appendix a , we prove that the global optimal solutions @xmath98 and @xmath99 are given by @xmath100 where @xmath101 is a real number . from the above result
, the optimal spreading sequence of the user @xmath9 , @xmath102 is given by @xmath103 for a real number @xmath101 .
this is equivalent to the following spreading sequences : @xmath104 where @xmath105 which satisfies @xmath106 when @xmath107 .
this sequence is obtained from the triple @xmath23 and @xmath108 in the fzc sequence families .
if @xmath109 , @xmath110 and @xmath111 , where @xmath112 is an even number , then our sequences are equivalent to the song - park ( sp ) sequences ( @xmath113 is not used ) @xcite .
if @xmath114 and @xmath115 , our sequences are equivalent to the sarwate sequences @xcite .
in the previous section , we fix the number of users @xmath43 and derive optimal spreading sequences in the weyl spreading sequence class .
this spreading sequence is expected to be useful when the number of the users @xmath43 is fixed . however , the number of users in a channel changes as time passes .
thus , in this section , we fix the maximum number of users in a channel , @xmath116 and assign the @xmath117 to @xmath43 users .
the spreading sequences assigned to the user @xmath9 is expressed as @xmath118 note that optimal spreading sequence of the problem ( @xmath119 ) is expressed as @xmath120 in particular , when @xmath121 and @xmath122 , this spreading sequence is equivalent to the sequence shown by sarwate @xcite .
this spreading sequence has the one that the periodic crosscorrelation function is always 0 , that is , @xmath123 for all @xmath124 .
note that @xmath125 in all the @xmath101 .
thus , we define @xmath116 as @xmath0 , that is , we consider the following sequences @xmath126 in this section , we assume that @xmath127 is a random variable and is uniformly distributed in @xmath128 . in the next section ,
we consider how to assign @xmath127 in a systematic approach .
the expression of snr of the user @xmath46 is obtained in @xcite @xcite as @xmath129 where @xmath130 \\ & + |c_{i , k}(l - n+1)|^2 + |c_{i , k}(l)|^2 \\ & + \operatorname{re}[c_{i , k}(l ) \overline{c_{i , k}(l+1 ) } ] + |c_{i , k}(l+1)|^2 \ } , \end{split } \label{eq : snr}\ ] ] @xmath131 is the energy per data bit and @xmath132 is the power of gaussian noise .
snr is the ratio between the variance of a desired signal and the one of a noise signal . in the appendix b , with spreading sequence @xmath133 , we prove that snr of the user @xmath46 is given by @xmath134 where @xmath135 equation ( [ eq : upper ] ) is obtained when the ratio @xmath136 is close to @xmath137 , that is , the number of users @xmath43 is sufficiently large . from eqs .
( [ eq : upper ] ) and ( [ eq : upper_sub ] ) , the spreading sequence @xmath138 has different snr in @xmath139 .
thus , some users have high snr and other users have low snr .
the lower bound of @xmath140 is @xmath141
in this section , we consider how to assign @xmath127 to the each users .
let us consider the spreading sequences @xmath142 from eq .
( [ eq : global_solution ] ) , it is demanded that we assign @xmath127 at regular interval .
however , these sequences can not be used if the number of users changes .
thus , we have to make the rule to assign @xmath127 when the number of users changes .
we give the rule to assign @xmath127 in the situation that the number of users monotonic increases . from the demanded of the problem @xmath80 ,
it is desirable that we assign @xmath127 to each users at regular interval .
thus , it is appropriate to assign them at nearly regular interval in every number of users .
we apply the van der corput sequence @xcite to the method to assign @xmath127 since the sequence is a regular interval sequence in some situations . for example
, the van der corput sequence @xmath143 is obtained as @xmath144 in particular , when we take the first eight elements out from @xmath143 and sort them , we obtain the sequence @xmath145 this sequence is a regular interval sequence . we can consider @xmath143 as a nearly regular interval sequence . when the length of spreading sequences @xmath0 equals @xmath146 , where @xmath147 is an integer , @xmath143 is rewritten in terms of @xmath148 .
for example , when @xmath149 , the sequence @xmath143 is obtained as @xmath150 thus , we propose that we use the @xmath9-th element of @xmath143 as @xmath151 , that is , the spreading sequences are expressed as @xmath152 where @xmath153 is the @xmath9-th element of @xmath143 .
in this section , we simulate an asynchrnous cdma communication system and discuss the performance of the spreading sequences obtained by eqs .
( [ eq : optimal_seq ] ) ( [ eq : sarwate_seq ] ) ( [ eq : vander_seq ] ) .
we use two parameters @xmath154 and @xmath155 , which are independent of the number of users @xmath43 and depend on @xmath43 .
we consider a bpsk model .
the channel has a additional white gaussian noise ( awgn ) and no fading signals .
we measure the average bit error rate that @xmath156 where @xmath157 is the trial numbers , @xmath158 is the @xmath158-th trial number and @xmath159 is the bit error rate of the user @xmath9 at the @xmath158-th trial . in each trial , we choose randomly the parameter the delay @xmath49 , the symbol @xmath160 and the phase @xmath161 and the initial elements of sequence @xmath127 .
this section consists of three subsections . in the first subsection , we compare the spreading sequences obtained by eq .
( [ eq : sarwate_seq ] ) and ( [ eq : optimal_seq ] ) with other sequences , the gold codes @xcite , the optimal chebyshev spreading sequences @xcite and the fzc sequence families @xcite .
in particular , we choose the triple @xmath162 as the parameters of the fzc sequence families .
this triple is shown in @xcite as the optimal parameters when @xmath163 with @xmath0 being the length . in the second subsection , we compare the spreading sequences obtained by eq .
( [ eq : sarwate_seq ] ) and ( [ eq : optimal_seq ] ) with one obtained by eq .
( [ eq : vander_seq ] ) .
we compare the random assigning approach with the systematic approach . in the final subsection , we compare the spreading sequences obtained by eq .
( [ eq : sarwate_seq ] ) and ( [ eq : optimal_seq ] ) with the sp sequences @xcite .
we consider the following spreading sequences : @xmath164 the former sequence is obtained from eq .
( [ eq : sarwate_seq ] ) and the latter sequence is obtained from eq .
( [ eq : optimal_seq ] ) . in this section ,
the `` weyl '' spreading sequence is @xmath165 and the `` optimal '' one is @xmath166 .
note that the optimal spreading sequences are different in the number of users @xmath43 .
the `` upper bound '' is obtained from eq .
( [ eq : upper ] ) .
figure 1 shows the relation between the number of users and bit error rate ( ber ) when @xmath163 and @xmath167(db ) . in this figure
, we set @xmath168 , which is independent of the number of users @xmath43 .
the ber of the weyl spreading sequences is lower than the one of the gold codes and the optimal chebyshev spreading sequences .
however , it is higher than one of the fzc sequence families .
the upper bound is established when the number of the users @xmath43 is larger than 20 . on the other hands ,
the ber of the global solution of the problem @xmath80 , the optimal weyl sequences @xmath166 is the lowest .
these sequences are dramatically efficient when the number of the users @xmath43 is fixed .
figure 2 shows the relation between the @xmath169(db ) and bit error rate ( ber ) when @xmath170 .
we used the two cases for @xmath101 , @xmath171 and @xmath172 .
the former parameter @xmath171 is independent of @xmath43 and the latter parameter @xmath172 is dependent on @xmath43 . in this figure ,
the ber of the weyl spreading sequences @xmath173 is lower than one of the fzc sequence families .
this result shows that ber of the weyl spreading sequences is changed when the value of @xmath101 is varied .
however , the ber of the optimal sequence , which is the global solution of @xmath80 , is independent of @xmath101 .
( db ) , @xmath0=31,width=288 ] ( db ) and ber : @xmath174 , @xmath0=31,width=288 ] in section v , we discussed how to assign the element @xmath127 to the user @xmath9 and proposed the method to assign .
we set the length @xmath175 and the parameter @xmath168 .
we compare two types of weyl spreading sequences , whose @xmath127 is randomly assigned to user @xmath9 and whose @xmath176 is orderly assigned as the @xmath9-th element of the van der corput sequence .
note that the first @xmath43 elements of the van der corput sequence are used as @xmath176 when the number of the users is @xmath43 .
figure 3 shows the relation between the number of users and ber when @xmath167(db ) .
the ber of the spreading sequences obtained by eq .
( [ eq : vander_seq ] ) is lower than one of the spreading sequences to which we randomly assign @xmath127 .
the weyl spreading sequences will have better performance if we successfully assign @xmath127 .
, width=288 ] the song - park ( sp ) sequences have been proposed in @xcite .
we set the length @xmath177 and the number of the users @xmath174 .
then , the maximum number of the users of the sp sequences is 14 .
thus , we compare them with four types of the weyl spreading sequences . we choose the parameters as @xmath178 .
the parameter @xmath179 represents the situation where the maximum number of the users is 14 .
thus , the weyl spreading sequences @xmath180 have the same feature to the sp sequences .
figure 4 shows the relation between the @xmath169(db ) and ber .
the ber of the weyl spreading sequences whose @xmath181 is higher than one of the sp sequences .
however , the ber of the weyl spreading sequences whose @xmath179 and one of the sp sequences are the same .
further , the ber of the weyl spreading sequences whose @xmath179 is the same ber in different @xmath101 .
this result and the result of subsection ( a ) could suggest that the optimal parameter @xmath101 depends on @xmath0 , @xmath43 and @xmath116 .
the ber of the global solutions is lowest and each ber of them is the same in @xmath171 and @xmath182 . by taking into account these ,
we conclude that the ber of the global solution is independent of @xmath101 .
( db ) and ber : @xmath174 , @xmath177,width=288 ]
in this paper , we have defined the weyl sequence class and shown the features of the sequences in the class .
we have constructed the optimization problem : minimize the upper bound of the absolute value of the whole interference noise and derive the global solutions . from this solution
, we can derive other sequences , the sarwate s sequences , the sp sequences .
the global solution is dramatically efficient when the number of the users @xmath43 is fixed .
we have evaluated their snr and shown the simulation results . in the global solution of the problem @xmath80 ,
the parameter @xmath101 is any real number .
however , its ber depends on @xmath101 when we let the maximum number of users @xmath116 as @xmath0 or other number , not @xmath43 .
the remained issue is to investigate the optimal @xmath101 and how to assign @xmath127 successfully to the user @xmath9 .
[ [ section ] ] in this appendix , we prove that the global optimal solutions of @xmath80 , @xmath98 and @xmath99 are given by @xmath183 where @xmath101 is a real number .
+ since the problem @xmath80 is a convex programming , it is necessary and sufficient for the global solution to satisfy the kkt conditions , eqs .
( [ eq : kkt1_1])-([eq : kkt1_3 ] ) .
+ when @xmath98 satisfies eq .
@xmath184 , it is clearly that @xmath185 since @xmath186 and @xmath187 .
we let @xmath188 .
thus , it is sufficient to consider only two kinds of the lagrange multipliers , @xmath189 and @xmath190 .
they satisfy the following equation which is obtained from eq .
( [ eq : kkt1_1 ] ) : @xmath191 where @xmath192 have @xmath137 in the @xmath46-th element and 0 in the others and @xmath193 have @xmath137 in the @xmath194-th element and 0 in the others . from eq .
( [ eq : kkt2 ] ) , we consider two vector equations .
one is the first @xmath43-dimensional vector equation of eq .
( [ eq : kkt2 ] ) and the other is the last @xmath195-dimensional vector equation . they are expressed as @xmath196 @xmath197 then , we define @xmath198 as @xmath199 note that @xmath200 since @xmath201 . from the definition of @xmath99 ,
@xmath198 only depends on the absolute value of difference , @xmath202 .
we therefore rewrite @xmath198 as @xmath203 the variable @xmath204 has the property that @xmath205 this result is obtained from the definition of @xmath99 .
we consider the two types of @xmath43 : @xmath43 is an odd number or @xmath43 is an even number .
for all @xmath46 and @xmath206 , @xmath98 , @xmath207 and @xmath99 satisfy either only @xmath208 or @xmath209 .
they satisfy @xmath210 we consider the @xmath14-th element of the left side of eq .
@xmath212 from eq .
@xmath213 , for all the integers
@xmath46 and @xmath9 , the term in summation of the left side of eq .
@xmath214 equals @xmath215 . from the above proof
, all the lagrange multipliers satisfy eq . @xmath184 .
the lagrange multipliers @xmath98 , @xmath207 and @xmath99 satisfy @xmath216 when @xmath217 , they satisfy @xmath218 and @xmath209 .
thus , we set @xmath219 similar to the case that @xmath43 is an odd number , we consider the @xmath14-th element of left side of eq . @xmath211 .
@xmath220 the terms of the difference equaling @xmath221 vanish .
therefore , we obtain @xmath222 thus , we have proven that eq .
( [ eq : la2 ] ) equals to 0 .
it is clearly that the left side of eq .
@xmath214 equals @xmath215 when @xmath223 .
when @xmath217 , it follows that @xmath224 thus , for all the integer @xmath46 and @xmath9 , eq .
@xmath214 is satisfied .
+ from the proofs a and b , we have proven that the existence of the lagrange multipliers which satisfy eq
. therefore , @xmath225 and @xmath99 are the global solutions of the problem ( @xmath119 ) .
[ [ section-1 ] ] in this appendix , with the spreading sequences @xmath133 , we prove that snr of the user @xmath46 is given by @xmath226 where @xmath227 we assume that the element @xmath127 is a random variable uniformly distributed in @xmath228 .
this assumption is fullfilled when the ratio @xmath136 is close to @xmath137 , that is , the number of users is sufficiently large since snr is not the reciprocal of the average of the interference noise over the users .
however , with the spreading sequences @xmath133 , they are equivalent when the number of users @xmath43 equals @xmath0 ( see eq .
( [ eq : first_r2 ] ) ) .
thus , it is conceivable that the assumption is established when the ratio @xmath136 is close to @xmath137 .
the correlation function @xmath75 of the spreading sequences in eq .
( [ eq : optimal_seq ] ) is @xmath229 where @xmath230 and @xmath231 thus , we obtain the squared absolute value of @xmath75 : @xmath232 on the other hand , the following relations are satisfied : @xmath233 @xmath234\\ = & \frac{n\left\{\cos\left(2 \pi \left(\gamma + \frac{\sigma_k}{n}\right)\right ) + \cos\left(2 \pi \left(\gamma + \frac{\sigma_i}{n}\right)\right)\right\}}{2\left(1-\cos\left(2 \pi \frac{\sigma_k - \sigma_i}{n}\right)\right ) } , \end{split}\ ] ] and @xmath235\\ = & \sum_{l=0}^{n-1}\operatorname{re}\left[c_{i , k}\left(l\right)\overline{c_{i , k}\left(l+1\right)}\right ] .
\label{eq : c_2 } \end{split}\ ] ] in the above equations , we used the assumption @xmath236 . from eqs .
( [ eq : c_1])-([eq : c_2 ] ) , @xmath237 in eq .
( [ eq : snr ] ) is given by @xmath238 when we calculate the sum of eq .
( [ eq : r_ik ] ) , the first term of it is given by @xmath239 the integer @xmath240 is a random variable and satisfies @xmath106 when @xmath107 .
thus , @xmath127 is expressed as @xmath241 and we can treat @xmath242 as a random variable instead of @xmath127 .
the integer @xmath242 is uniformly distributed in @xmath243 .
thus , the average of eq .
( [ eq : first_r ] ) is @xmath244 where @xmath245 is the average over @xmath127 . in @xcite
, it is shown that @xmath246 thus , eq . ( [ eq : first_r2 ] ) is equivalent to @xmath247 from the above result , we obtain the following relation @xmath248 the average of the second term of the sum of eq .
@xmath249 is given by @xmath250 note that it is clear that @xmath251 therefore , eq . ( [ eq : term2 ] )
is rewritten as @xmath252 thus , the sum of the average of the second term in eq .
( [ eq : r_ik ] ) is written as @xmath253 similarly , we obtain the average of the sum of the third term of eq .
( [ eq : r_ik ] ) : @xmath254 finally , we obtain the following relation @xmath255 from eq . @xmath256 and eq .
@xmath257 , we arrive at snr of the user @xmath46 with the spreading sequence @xmath102 @xmath226 where @xmath258
one of the authors , hirofumi tsuda , would like to thank for advise of prof .
nobuo yamashita and dr .
shin - itiro goto .
99 c. e. shannon , `` a mathematical theory of communication , '' _ bell system technical journal _ ,
volume 27 , issue 3 , 379 - 423 ( 1948 ) .
s. verd and s. shamai , `` spectral efficiency of cdma with random spreading . ''
_ ieee transactions on information theory _ 45.2 ( 1999 ) : 622 - 640
. j. proakis , `` digital communications .
1995 '' , mcgraw - hill , new york .
r. steele and l. hanzo , `` mobile radio communications '' , second and third generation cellular and watm systems : 2nd .
ieee press - john wiley , 1999 .
m. honig , u. madhow and s. verdu , `` blind adaptive multiuser detection . ''
_ ieee transactions on information theory _ 41.4 ( 1995 ) : 944 - 960 .
t. ristaniemi and j. joutsensalo , `` advanced ica - based receivers for block fading ds - cdma channels '' , _ signal processing _ , 82.3 ( 2002 ) : 417 - 431 .
a. hyvrinen , j. karhunen , and e. oja , `` independent component analysis '' , vol .
john wiley @xmath259 sons , 2004 .
s. verdu , `` minimum probability of error for asynchronous gaussian multiple - access channels '' , _ ieee transactions on information theory _ , 32.1 ( 1986 ) : 85 - 96 .
r. gold , `` optimal binary sequences for spread spectrum multiplexing ( corresp . ) . ''
_ ieee transactions on information theory _ , 13.4 ( 1967 ) : 619 - 621 . t. kasami , `` weight distribution formula for some class of cyclic codes . ''
_ coordinated science laboratory report _
r-285 ( 1966 ) .
g. heidari - bateni and c. d. mcgillem , `` a chaotic direct - sequence spread - spectrum communication system . ''
_ ieee transactions on communications _ ,
42.234 ( 1994 ) : 1524 - 1527 .
halle , c.w .
wu , m. itoh and l.o .
chua , `` spread spectrum communication through modulation of chaos . '' _ international journal of bifurcation and chaos _ , 3.02 ( 1993 ) : 469 - 477
. y. soobul , k. chady and h. c.s .
rughooputh , `` digital chaotic coding and modulation in cdma . ''
_ africon conference in africa , 2002 .
ieee africon .
6th_. vol .
2 . ieee , 2002 .
chen , k. yao , k. umeno and e. biglieri , `` design of spread - spectrum sequences using chaotic dynamical systems and ergodic theory . ''
_ ieee transactions on circuits and systems i : fundamental theory and applications _ , 48.9 ( 2001 ) : 1110 - 1114 .
k. umeno and k. kitayama .
`` spreading sequences using periodic orbits of chaos for cdma . ''
_ electronics letters _ 35.7 ( 1999 ) : 545 - 546
. g. mazzini , r. rovatti and g. setti , `` interference minimisation by autocorrelation shaping in asynchronous ds - cdma systems : chaos - based spreading is nearly optimal . ''
_ electronics letters _ , 35.13 ( 1999 ) : 1054 - 1055 . g. mazzini , g. setti , and r. rovatti , `` chaotic complex spreading sequences for asynchronous ds - cdma .
i. system modeling and results . ''
_ ieee transactions oncircuits and systems i : fundamental theory and applications _ ,
44.10 ( 1997 ) : 937 - 947 .
r. riccardo , g. mazzini and g. setti , `` on the ultimate limits of chaos - based asynchronous ds - cdma - i : basic definitions and results '' , _ ieee transactions on circuits and systems _
i , 51.7 ( 2004 ) : 1336 - 1347 .
d. v. sarwate , `` bounds on crosscorrelation and autocorrelation of sequences '' , _ ieee transactions on information theory _ , 25.6 ( 1979 ) : 720 - 724
. l. r. welch , `` lower bounds on the maximum cross correlation of signals '' , _ ieee transactions on information theory _ , 20.3 ( 1974 ) : 397 - 399
. r. frank , s. zadoff and r. heimiller , `` phase shift pulse codes with good periodic correlation properties ( corresp . ) . '' _ ire transactions on information theory _ 8.6 ( 1962 ) : 381 - 382
. d. chu , `` polyphase codes with good periodic correlation properties ( corresp . ) . ''
_ ieee transactions on information theory _ 18.4 ( 1972 ) : 531 - 532 .
j. oppermann and b. s. vucetic , `` complex spreading sequences with a wide range of correlation properties . ''
_ ieee transactions on communications _ , 45.3 ( 1997 ) : 365 - 375
. h. weyl .
ber die gleichverteilung von zahlen mod .
ann , 77:313 - 352 , 1916 .
( in german ) . j. dick and f. pillichshammer , `` digital nets and sequences : discrepancy theory and quasi - monte carlo integration '' . cambridge university press , 2010 . k. umeno , `` spread spectrum communications based on almost periodic functions '' _ ieice technical report _
, nlp 2014 - 101 , pp .
11 - 16 ( 2014)(in japanese ) d. v. sarwate and m. b. pursley , `` crosscorrelation properties of pseudorandom and related sequences . ''
_ proceedings of the ieee _ , 68.5 ( 1980 ) : 593 - 619 . m. b. pursley , `` performance evaluation for phase - coded spread - spectrum multiple - access communication .
i - system analysis . '' _ ieee transactions on communications _ , 25 ( 1977 ) : 795 - 799
. w. kuhn and a. w. tucker , `` nonlinear programming '' , in j. neyman ( ed . ) , proceedings of the second berkley symposium on mathematical statistics and probability ( university of california press , berkley , ca ) , pp .
481 - 492 , 1951 .
s. r. park , l. song and s. yoon , `` a new polyphase sequence with perfect even and good odd cross - correlation functions for ds / cdma systems . '' _ ieee transactions on vehicular technology _
, 51.5 ( 2002 ) : 855 - 866 . j.g .
van der corput , `` verteilungsfunktionen i und ii '' .
wet_. , 38 ( 1935 ) , p. 813 .
e. r. hansen , `` a table of series and products . ''
prentice hall series in automatic computation , englewood cliffs : prentice hall , 1975 1 ( 1975 ) , eq .
( 24.1.2 ) . | this paper shows an optimal spreading sequence in the weyl sequence class belonging to the extended frank - zadoff - chu ( fzc ) sequence family .
the sequences in this class have the desired property that the order of crosscorrelation is low .
we evaluate the upper bound of crosscorrelation in the class and construct the optimization problem .
then , we derive the optimal spreading sequences as the global solution of the problem and show their snr in a special case . from this result , we propose how the initial elements are assigned , that is , how spreading sequences are assigned to each users .
we also numerically compare our spreading sequences with other ones , the gold codes , the sequences of the fzc sequence families , the optimal chebyshev spreading sequences and the sp sequences in bit error rate .
submitted paper spread spectrum communication , nonlinear programming , spreading sequence , signal to noise ratio , bit error rate |
radiation matter interaction has been studied for a long time with atomic samples .
it is known that density matrix theory is well suited to exhaustively describe the properties of fluorescence from macroscopic atomic ensembles , and that in this limit semiclassical theory gives excellent approximations , since phenomena related to the quantum nature of the light field are often hidden . on the other hand ,
it has also been realized since more than two decades now that isolation of a single atom provides an opportunity to directly observe pure quantum properties of the interacting radiation - matter system .
a celebrated example is the observation of photon antibunching , where initial experiments @xcite were carried out with extremely diluted atomic beams , and more precise investigations became possible when ion traps were employed to achieve long - term confinement of a single atomic particle @xcite . for neutral atoms
the confinement strength is reduced since trapping has to rely on forces derived from electric or magnetic dipole interaction .
therefore neutral atoms have only more recently become available for measurements at the microscopic level through the application of laser cooling techniques well established for macroscopic samples with many atoms .
with laser light single neutral atoms can be trapped and observed for long times @xcite in analogy with ion traps .
one of the most powerful methods to observe quantum fluctuations in atom - radiation interaction is to measure photon correlations of the light emitted by a single atom .
the quantum theoretical description of photodetection was put forward by glauber @xcite in the early 60s , and very successfully applied to all appropriate experiments since . in the photon language
the normalized intensity - intensity correlation function is the conditional probability to detect a second photon if a first one was detected a time @xmath0 before .
this second order correlation function is defined by @xmath1 where @xmath2 is the photon number operator constructed from field operators @xmath3 , @xmath4 , and where @xmath5 denotes normal ordering of field operators . at sufficiently long delays
@xmath0 all possible correlations have decayed , and hence @xmath6 .
it is thus often informative to discuss the deviation of @xmath7 from unity , or the quantity @xmath8 as a measure of the fluctuation strength of the system , called _
contrast _ in the following .
for auto correlations ( [ eq : g2 ] ) all classical fields must obey @xmath9 , while photon antibunching shows @xmath10 , making it a prime example of quantum field fluctuations . in an intuitive interpretation it is said @xcite that the first detected photon projects an atom in its ground state . from this initial state
the atom then relaxes back to its equilibrium state on the time scale of its radiative decay . in this sense
the observation of this fluctuation is measurement induced. in our work we are additionally interested in the case where different atomic transitions can be distinguished by their polarization properties .
we therefore analyze cross correlations of orthogonal polarization states @xmath11 and @xmath12 of the resonance fluorescence from a single atom , @xmath13 a single trapped cesium atom shows very strong polarization correlations in its resonance fluorescence .
while this effect is very pronounced for correlations between orthogonal circular polarizations of the detected photons , it vanishes for linearly polarized photons even in a light field with linear polarization at every place .
it is the purpose of this work to show that the strong contrast in @xmath14 that we observe for orthogonally circular polarization states in the fluorescence of a single trapped atom can be interpreted as a direct consequence of the atomic orientation or magnetization undergoing spontaneous or quantum fluctuations .
all classical fields have auto correlations @xmath15 , and a value @xmath16 is classically forbidden ( for classical fields one should replace all operators @xmath17 in ( [ eq : g2 ] ) by their classical counterparts , intensities @xmath18 ) .
this enhanced probability to detect two photons simultaneously is called photon bunching and was observed as early as in 1956 @xcite from a usual thermal light source . for a single - atom fluorescence , however , @xmath19 vanishes identically for @xmath20 which is a reflection of the fact that one can find at most one photon in the field mode of interest and can never detect two photons simultaneously .
this phenomenon is called photon antibunching and is regarded as an important manifestation of the quantum nature of light .
since the emitted light field reflects the evolution of the atomic dipole moment the correlation function @xmath19 visualizes the internal dynamics of the observed atom for @xmath21 .
the state of an excited atom evolves continuously in the absence of a measurement , but theory predicts a sudden projection to the ground state when a photon is detected . this measurement triggers the atom to the initial conditions @xmath22 and @xmath23 , where @xmath24 and @xmath25 represent the population of the ground and excited atomic state , respectively . at that instant
the coherent evolution starts again from the values @xmath22 and @xmath23 and will be interrupted by the next spontaneous emission .
the normalized probability for detecting a second spontaneously emitted photon is now proportional to the population of the excited atomic state @xmath26 according to @xcite @xmath27 since the emission times are random , after averaging over many evolution trajectories the measured @xmath7 shows directly relaxation of the system back to the equilibrium state after its wave function has collapsed due to the first photon detection . in fig .
[ fig : rabi1 ] we show @xmath7 of resonance fluorescence from a single atom stored in a magneto - optical trap ( mot ) @xcite with clearly observable photon antibunching and transient oscillations . as we already reported previously @xcite , the observed transient oscillations in the population of the excited state in fig .
[ fig : rabi1 ] corresponding to coherent excitation and deexcitation cycles ( rabi oscillations ) can be surprisingly well described by a simple model ( solid line in fig . [ fig : rabi1 ] ) of a two - level atom ( similar to @xmath28 transition ) in spite of the complicated multilevel structure of the cesium atom and light interference pattern ( see below ) .
this observation suggests that due to optical pumping a trapped atom spends most of its time in the magnetic substate that interacts most strongly with the local field and is forced to behave , to a good approximation , like a two - level system . for @xmath0 larger than the life time of the excited state the correlations die out due to the fluctuations of the vacuum field .
an atom trapped in a mot and moving through the light interference pattern experiences various intensities and polarizations at different places .
the polarization of the resonance fluorescence is determined by the atomic interaction with the local light field and changes on the time scale of atomic transport over an optical wavelength @xmath29 .
thus , in addition to correlations of the total intensity one expects also polarization effects , that is correlations @xmath14 measured between any polarization components @xmath30 and @xmath31 which should strongly depend on the atomic motion and the light - field topography .
the light field of the mot is formed by three mutually orthogonal pairs of counterpropagating laser beams with @xmath32 and @xmath33 polarization .
a pair of two circularly polarized laser beams with the same handedness produces a local polarization that is linear everywhere with a direction of polarization that rotates a full turn every half wavelength . with two additional polarization screws for other directions along with the relative phases @xmath34 and @xmath35 between these standing waves one obtains for the total electric field in a 3d mot@xmath36 @xmath37 we have already reported on strong correlations between circularly polarized photons and vanishing correlations between linear polarization components observed in the resonance fluorescence of a single atom trapped in a standard mot light field configuration @xcite .
this result seems to be intuitive only for specific choices of time phases : for example , @xmath38 yields an antiferromagnetic light - field structure with alternating right- and left - hand circular polarizations at points of deepest light shift potential , see @xcite .
however , in a standard mot the phases @xmath34 and @xmath35 change randomly due to acoustic jitter and thermal drifts .
thus the trapping light field has no well defined polarization state and the correlations are averaged over all possible values of @xmath34 and @xmath35 . since the mot light field topography ( [ eq:3dmot ] )
strongly depends on the relative time - phases of the three contributing standing waves , we have chosen a setup where @xmath34 and @xmath35 are intrinsically stable @xcite .
the concept uses a single standing wave which is multiply folded and brought into triple intersection with itself .
the phases @xmath34 and @xmath35 can be adjusted by means of faraday rotators .
details of this approach have been published elsewhere @xcite .
if one models an atom by a `` classical emitter '' ( @xmath28 transition or steady - state density matrix of a multi - level atom ) then its induced dipole moment will be proportional to the local light field .
this idea is often used for interpretation of polarization correlations in the fluorescence of a large number of laser - cooled atoms @xcite .
as we will show this description fails completely in the case of a single atom .
the most interesting case occurs for @xmath39 .
in this situation the three standing waves oscillate synchronously and thus the interference light field has a linear polarization at every point and lacks handednees completely . in this case the model of a classical emitter @xcite predicts strong correlations between orthogonal linear polarization components and relatively weak correlations between circular components in clear contradiction with experimental results ( see below ) .
for polarization - sensitive correlation measurements we have trapped individual neutral atoms in a standard six - beam mot @xcite with the only exception that the phases and hence the light field topography is fully controlled @xcite . at a quadrupole field gradient of 12.5
g / cm the storage volume extends over approximately 100 @xmath40 m . in order to trap small , countable numbers of atoms the loading rate from the background atomic vapour into the trap
is kept very low .
this is achieved on the one hand by lowering the cesium partial pressure to @xmath41 mbar ( at a base pressure of @xmath42 mbar ) and on the other hand by using trapping laser beams of diameter 4 mm only .
the average number @xmath43 of trapped atoms ( typically between 1 and 5 in this experiment ) can be easily adjusted by variation of the cesium pressure .
although @xmath43 also depends on the trapping laser intensity @xmath44 and detuning @xmath45 of the trapping laser from atomic resonance , we are able to observe trapping of individual atoms over a wide range of parameters : @xmath46 per laser beam and @xmath47 .
the natural linewidth and the saturation intensity of the cooling transition are @xmath48 mhz and @xmath49 mw/@xmath50 , respectively .
the atomic resonance fluorescence is due to excitation by the trapping laser field only .
fluorescent light is collected from a @xmath51 solid angle by a lens and then splitted into orthogonal polarization states by means of polarizing optics ( see fig .
[ fig : poloptics ] ) which also directs the corresponding light onto two avalanche photo diodes ( apd ) .
the apds are operated in single photon counting mode and achieve a photon detection efficiency of 47 % at a dark count rate of 10 s@xmath52 . the average photon count rate for an individual atom lies in the range 3 - 10 khz in our experiments , depending on laser intensity and detuning .
observation direction is in the xy - plane at 45@xmath53 to the laser beams ( the @xmath54-axis is the symmetry axis of the mot quadrupole magnetic field ) .
usually , a measurement of the cross correlation function along with the total intensity - intensity correlation provides complete information : the corresponding auto correlation can be inferred from the sum rules for orthogonal polarization components @xcite .
for example for circular components one has @xmath55 .
a cross correlation measurement ( @xmath56 ) can be carried out 4 times faster than the corresponding auto correlation measurement ( @xmath57 ) where only one half of the total fluorescence is detected .
a computer registers the arrival times of all photons from the two apd channels with 100 ns time resolution and with 700 ns dead time in each channel .
the entire experimental information accessible in the set - up is stored and thus can be processed afterwards by correlation analysis through numerical multi - stop procedures , which completely eliminate systematic errors such as photon pile - up introduced by single - stop methods traditionally used in hanbury brown & twiss type experiments @xcite .
atoms are randomly loaded from background vapour and randomly lost due to collisions with background gas .
but since individual atom arrival and departure events are easily located within 1 ms , it is straightforward to determine the instantaneous number of atoms from the average count rate .
note that the number of trapped atoms fluctuates on the second time scale @xcite .
this enables us to separate all data from a single experimental run into different classes with the number of trapped atoms as a parameter - the data for different atom numbers are therefore obtained under identical experimental conditions .
it is also easily possible to distinguish correlations of the fluorescence of trapped individual atoms from uncorrelated background of detection events due to stray light or fluorescence from thermal , untrapped atoms . as a consequence
all measured correlation functions can always be unequivocal normalized .
measured second order correlation functions for orthogonal polarization components in the fluorescence of a single atom in the mot@xmath58 are shown in fig .
[ fig : pol_corr ] . within our experimental uncertainties correlations
are completely absent for linear polarization components of the fluorescence , in baffling contrast with the result for circularly polarized components . for a single atom the circular correlation contrast reaches s values
up to 62% .
this very strong correlation typically relaxes within a few @xmath40s .
since photon antibunching and rabi oscillations at ns time scales are not resolved in these measurements , we have assumed for a simple analysis the relaxation process of the polarization correlations to be exponential and extracted a single relaxation time constant @xmath59 .
the intensity - intensity auto correlation @xmath7 also shows a contrast of about 30% due to intensity modulations of the mot light field , but the relaxation time constant of about 0.6 @xmath40s in fig .
[ fig : int_corr ] is significantly shorter than for the circular cross correlations @xcite .
furthermore , we have experimentally verified that @xmath59 does not depend on the number of trapped atoms and that the contrast of the correlation function is proportional to the inverse number of atoms , @xmath60 ( see fig .
[ fig : ndepend ] ) .
thus , as one would expect , we deal with a pure single - atom effect .
in order to interpret the polarization properties of the resonance fluorescence of atoms driven by a light field with linear polarization we begin by considering an atom at rest . as this light field
does not favor any of the two circular polarization states , one can assume that prior to the detection of the first photon the distribution of atomic magnetic sublevels is symmetric , @xmath61 .
if we suppose for simplicity that the local light field consists of equal parts of both orthogonal circular polarizations only , it is clear that the @xmath61 state is unstable @xcite . at the level of an individual particle this equilibrium state
can be distorted by the observation of a single circularly polarized photon , which projects the atom into its ground state , breaks the symmetry of the zeeman substate population ( fig .
[ fig : pops ] ) , and creates an imbalance in the interaction strengths with both circular polarization components . the next absorption will be preferentially further enhance the asymmetry .
the imbalance in the interaction strengths rapidly grows with @xmath62 leading to fast pumping into one of the outmost zeeman states @xmath63 .
the ratio of the interaction strengths in these stretched states reaches the value @xmath64 making them very stable for large @xmath65 even in the presence of the polarization component in the light field driving @xmath66 transitions .
the atom in this oriented state prefers to radiate into the same polarization state as the first detected photon , resulting in anticoincidences in the cross correlation for orthogonal circular polarizations .
note that this effect is a specific feature of @xmath67 transitions and does not occur for @xmath68 or @xmath69 transitions .
it can be regarded as bistability , since an emission of a circularly polarized photon makes the atom more likely to be pumped in the corresponding outmost zeeman state . as both stretched states @xmath70 and @xmath71 are equivalent in the presence of a linear polarized light field , this effect is principally unobservable in an atomic ensemble . however ,
if one adds some circular polarization component to the driving light field a spontaneous spin polarization of a macroscopic sample can be observed as recently demonstrated in @xcite by optically pumping on the @xmath72 hyperfine component of the @xmath73-line of cs .
although the real 3d - situation in our experiment is much more complicated than the simple model presented above , our interpretation is furthermore supported by the following considerations : as noted in @xcite , the condition for magnetic bistability in a linearly polarized light field is given by @xmath74 , where @xmath31 is the angle between the magnetic field ( quantization axis ) and the light polarization . in the mot@xmath58
there are 8 intensity antinodes in a unit cell with directions of the local linear polarization at these points coinciding with the diagonals of the coordinate system @xmath75 .
it is easy to see that the bistability condition is fulfilled for the majority of points in the mot magnetic quadrupole field @xmath76 .
however , in our case the stretched states are not intrinsically stable .
while radiation pressure forces are balanced for an aligned atom with @xmath77=0 , they are unbalanced for an oriented atom since the local linearly polarized light field is created by ( at least two ) _ counterpropagating _ laser beams with orthogonal circular polarizations .
the imbalance in the light forces created by atomic orientation thus causes acceleration , or heating which is again damped by the usual laser friction forces @xcite .
thus for our experiment we must acknowledge that the observation of a circularly polarized photon not only redefines atomic orientation but also its mechanical status : internal and external atomic degrees of freedom are inextricably entangled . the measured relaxation time constant of @xmath78 indeed depends strongly on the atom - light field interaction which also governs atomic motion in the trap .
the interaction strength is measured by the light shift parameter @xmath79 @xcite corresponding to the maximum energy shift of the atomic energy levels . under our experimental parameters
sub - doppler cooling leads to atomic temperatures proportional to the light shift , thus @xmath80 @xcite . in fig.[fig : tauvslsh ] we show the measured dependence of the time constant @xmath59 as a function of @xmath79 .
as expected we find good agreement of our experimental data with the functional relationship @xmath81 . here
@xmath82 denotes the average atomic velocity .
this means that relaxation of the spontaneous atomic magnetization is determined by atomic motion through the light field .
we can carry out our analysis one step further if we assume that the characteristic length over which relaxation takes place is @xmath83 = @xmath84 , the spatial period of the mot light field .
it is then straightforward to evaluate characteristic kinetic temperatures from our correlation measurements to be in the range between 10 - 68 @xmath40k for the @xmath79 range shown in fig .
[ fig : tauvslsh ] , in good agreement with previous measurements @xcite .
photon correlations observed in the resonance fluorescence of a single atom provide a direct access to the internal atomic dynamics .
we have shown two examples of resolved quantum dynamics of an isolated atom stored in a magneto - optical trap . beyond observation of the well - known phenomenon of transient rabi oscillations ( usually connected with photon antibunching ) ,
we have observed fluctuations of the atomic magnetic orientation by measuring photon correlations between orthogonal polarization components . using a simple model we have given evidence for the following dynamical processes causing strong circular cross correlations in resonance fluorescence : spontaneous emission of circularly polarized photons causes instantaneous orientation .
subsequent photons are preferentially absorbed and emitted with identical polarization .
this memory effect leading to correlated absorption of photons with equal polarization and thus to increased momentum diffusion of an atom with a multi - level structure has been discussed in @xcite . in our experiments
we can clearly isolate this effect by observation of an anticorrelation of circularly polarized photons successively emitted with opposite handedness .
subsequent optical pumping induced by atomic motion in the light field causes relaxation of the orientation clearly seen in the photon correlations . in a sense
we have seen the elementary sub - doppler cooling and heating forces in a @xmath85-molasses at work .
we thank svenja knappe for providing some basic techniques in an early stage of the experiment .
we are also pleased to acknowledge frans e. van dorsselaer and gerhard nienhuis for fruitful discussions and sharing their insight into elementary atomic processes .
this work is supported by the deutsche forschungsgemeinschaft ( dfg ) .
rateike , g. leuchs , and h. walther , results cited by j.d .
cresser , _ et al .
_ , in _ dissipative systems in quantum optics _ , edited by r. bonifatio , topics in current physics , vol .
27 ( springer , berlin , 1982 ) , p. 21 .
r.j.glauber , the quntum theory of optical coherence , phys . rev .
* 130 * , 2529 ( 1963 ) ; r.j.glauber , coherent and incoherent states of the radiation fields , phys
. rev . * 131 * , 2766 ( 1963 ) phys .
131 , 2766 ( 1963 ) e.l .
raab , m. prentiss , a. cable , s. chu , and d.e .
dritchard , trapping of neutral sodium atoms with radiation pressure , phys .
rev . lett.*59 * , 2631 ( 1987 ) ; for a recent review see c.g .
townsend , n.h .
edwards , c.j .
cooper , k.p .
zetie , and c.j .
foot , a.m. steane , p. szriftgiser , h. perrin , and d. dalibard , phase - space density in the magneto - optical trap , phys .
rev.*a52 * , 1423 ( 1995 ) c. jurczak , b. desruelle , k. sengstock , j .- y .
courtois , c. i. westbrook , and a. aspect , atomic transport in an optical lattice : an investigation through polarization - selective intensity correlations , phys .
* 77 * , 1727 ( 1996 ) the difference in time constants for @xmath86 and @xmath7 rapidly growing with @xmath65 has been also observed in quantum monte - carlo simulations performed for an atom with @xmath67 transition in a 1d-@xmath85-molasses by f. e. van dorsselaer and g. nienhuis ( private communication ) .
similar effect of the growing difference in the cooling and magnetization decay times with increasing angular momentum has been recently mentioned in g. raithel , w. d. phillips , and s. l. rolston , magnetization and spin - flip dynamics of atoms in optical lattices , phys .
rev . * a 58 * , r2660 , ( 1998 ) | isolation of a single atomic particle and monitoring its resonance fluorescence is a powerful tool for studies of quantum effects in radiation - matter interaction . here
we present observations of quantum dynamics of an isolated neutral atom stored in a magneto - optical trap . by means of photon correlations in the atom s resonance fluorescence we demonstrate the well - known phenomenon of photon antibunching which corresponds to transient rabi oscillations in the atom . through polarization - sensitive photon correlations
we show a novel example of resolved quantum fluctuations : spontaneous magnetic orientation of an atom .
these effects can only be observed with a single atom .
keywords : laser cooling and trapping , single atoms , photon correlations , quantum fluctuations |
the study of the interactions between the quark and gluon constituents of hadrons has evolved over the past several decades .
experiments once had large systematic and statistical uncertainties and theoretical predictions used only leading - order perturbation theory .
however our understanding has considerably improved and precise measurements and calculations are now available .
the theory of the strong interactions , called quantum chromodynamics or qcd @xcite , is a very interesting one in its own right . in addition , because these kinds of interactions dominate at a hadron collider , they form a substantial background for other interesting possible physical processes , including top quark and higgs boson production , as well as other potential new pheneomena , such as models incorporating the principle of supersymmetry . in these proceedings
, we present recent qcd results , focusing predominantly on data taken at the fermilab tevatron .
these subjects include simple jet production , as well as jet production in association with electroweak bosons .
the fermilab complex accelerates protons and antiprotons and collides them in their joint center of mass frame at an energy of 1.96 tev .
these collisions are recorded by the d @xcite and cdf @xcite detectors , two large , multi - purpose , detectors located at the tevatron .
each detector consists of an inner tracker , composed of a silicon vertex detector and a detector of coarser granularity .
the tracking volume is surrounded by a superconducting solenoid and calorimetry .
the entire detector is enclosed by a magnetized muon tracking detector .
the instantaneous luminosity is as much as @xmath2 @xmath3 s@xmath4 and each experiment has recorded approximately 6 fb@xmath4 .
the various results presented here make use of 0.3 - 2.5 fb@xmath4 of data .
high energy jets are the debris of hadron interactions , which are often modelled as the hard scatter of partons from within the hadrons . in order to compare experimental measurements ( which involve many particles ) to theoretical calculations ( which generally involve very few ) ,
an algorithm is necessary that ( a ) integrates theoretically - intractable phenomena and ( b ) is valid over at all levels : parton , particle and detector .
there are two classes of jet - finding algorithms that are used , the cone - based algorithm and some sort of recombination algorithm . in this proceedings ,
the results mostly use a cone - based algorithm @xcite which iteratively combined energy with in a cone of radius @xmath5 where @xmath6 is the azimuthal angle , @xmath7 $ ] is the rapidity , and @xmath8 is the direction of the proton beam .
depending on the analysis , a cone size of @xmath9 or @xmath10 is used .
the seeds for the algorithm were provided by towers in the calorimeter or individual particles or partons . in order to stabilize the algorithm against low energy emission ,
the midpoint between reconstructed jets provided an additional set of seeds .
a second algorithm used in these proceedings is the @xmath11 algorithm @xcite .
this successive recombination algorithm uses all pairs of seeds to calculate @xmath12 , with @xmath13 in this proceedings and the @xmath14 is the transverse momentum of the @xmath15 seed .
this is then ordered in value and compared to the list of the transverse momentum squared of the remaining seeds ( @xmath16 ) .
if the minimum is a @xmath17 , this it is declared to be a jet and removed from further consideration .
if the minimum is one of the @xmath18 , the two are combined and the process repeated .
this algorithm is intrinsically safe to infrared emission .
cdf has published the inclusive jet cross section for both the cone @xcite and @xmath11 @xcite algorithms .
both analyses present the data in five bins in rapidity , covering a range to @xmath20 .
the results include approximately 1 fb@xmath4 of data .
the measurement was corrected to the particle level , as were the theoretical calculations .
the systematic uncertainties are substantial and are dominated by the jet energy scale correction .
figures [ figure_cdf_inclusive_cone ] shows the ratio of the cone measurement to theory , while figure [ figure_cdf_inclusive_kt ] shows the corresponding measurement using the @xmath11 algorithm .
inclusive spectrum.,width=302 ] the d experiment has published @xcite a measurement of the inclusive jet cross section as a function of jet @xmath19 in six rapidity bins , covering a range of @xmath21 . exploiting the liquid argon / uranium calorimetry , along with a detailed understanding of the response of the calorimeter to both quarks and gluons
, they were able to measure these quantities with unprecedented precision ; approximately 30 - 50% smaller than comparable cdf measurements .
figure [ figure_d0_inclusive_cone ] shows the ratio of the measurement to next - to - leading order theoretical calculations , using the cteq 6.5 m structure functions .
the data clearly is below the calculations at high jet @xmath19 .
this observation is taken as evidence that the pdfs used in this analysis might have too much momentum carried by gluons with a large fraction of the beam momenta .
this data was included in a recent pdf extraction @xcite . both d and cdf have published analyses studying the dijet mass spectrum , based on the same data set as was used for the inclusive jet analysis . both experiments limited themselves only to measurements using the cone algorithm .
the cdf measurement @xcite was restricted to a single rapidity range extending to @xmath22 , while d s measurement @xcite consisted of six bins in rapidity , extending to @xmath21 .
figures [ figure_d0_dijet_mass ] and [ figure_cdf_dijet_mass ] tell a story comparable to the jet inclusive @xmath19 measurements .
however , for this analysis , the d measurement compares to theory utilizing the mstw2008 pdfs .
these pdfs include tevatron run ii data ( including the d inclusive jet cross section which is based on the same data set ) , but no tevatron run i data .
in addition , cdf has used its dijet mass distribution ( shown in figure [ figure_cdf_dijet_resonance ] ) to search for dijet resonances .
they have set limits on axigluon / coloron , e6 diquark , color octet ( techni-@xmath23 ) and excited quarks .
table [ table_cdf_resonance_limits ] shows the published limits and a detailed description of the various models can be found in the references .
.cdf : limits set on new physics using their dijet mass resonance search . [ cols="^,^ " , ] both d @xcite and cdf @xcite have published analyses studying the dijet angular distribution as a function of the dijet mass . this particular analysis is attractive in that it is sensitive to searches for new physics and is relatively insensitive to uncertainties in the pdfs .
the essence of this approach is that standard qcd scattering occurs at small angles with respect to the beam direction , while new phenomena is expected to occur at larger angles .
the variable used in this analysis is @xmath24 where @xmath25 and @xmath26 are the rapidities of the two leading jets , ordered in jet @xmath19 .
figure [ figure_d0_angular ] shows d s published result of the normalized @xmath27 distribution for eight bins , including the first data in this variable for a dijet mass exceeding 1.1 tev .
the data is compared to several models , including quark compositeness , and two models of extra dimensions .
the data agrees well with the standard model and stringent limits are set on these speculative processes . for quark compositeness
, limits can be set on the compositeness scale , which must exceed 2.91(2.97 ) tev for constructive(destructive ) interference .
figure [ figure_cdf_angular ] shows cdf s ratio of data to monte carlo predictions for the integral of the @xmath27 distribution for @xmath28 .
the data is in good agreement with qcd prediction , with no hint of physics outside the current model .
distribution for @xmath28 for data over monte carlo predictions.,width=302 ]
while the production of jets has historically been the main focus of qcd analyses , events which contain both jets and vector bosons are an ideal sample to investigate physical processes with relatively low cross - sections .
the existence of the vector boson greatly suppresses the backgrounds .
typical physical processes of interest include top quark production , higgs boson production and various new phenomenon models . both d and cdf have made extensive measurements of events containing vector bosons and up to four jets .
cdf has reported @xcite a study of @xmath29 jet production using 320 pb@xmath4 of data where @xmath30 is up to 4 .
this study used the electron decay mode of the @xmath0 boson .
the jets found in this analysis used a cone - based algorithm with @xmath31 .
it required the electron to have @xmath32 gev and @xmath33 .
the neutrino was required to have @xmath34 gev and the transverse mass of the @xmath0 was @xmath35 gev .
the data was compared to both nlo and lo calculations , with the details found in the reference .
as approximately expected , data was found to be in good agreement with nlo and to have good shape agreement with lo .
leading order calculations appear to under - predict the data by about 40% .
distribution in @xmath0 + jet events for up to three jets.,width=302 ] cdf also has a published an analysis @xcite describing the behavior of events which have a @xmath1 boson and 1 - 3 jets .
the paper compares the data to nlo theory and good agreement is found both in the jet multiplicity distribution and in the jet @xmath19 distribution ( shown in figure [ figure_cdf_z_jets ] ) .
the systematic uncertainties are on the order of 10% .
leading order calculations have the right shape , but the wrong normalization .
distribution for @xmath1 + jet events , for the two leading jets .
in addition , a ratio is formed with nlo theory.,width=302 ] d has published analyses of @xmath1 + @xmath30 jet events , with both electron @xcite and muon @xcite decay channels .
the data is compared to both lo and nlo calculations @xcite and leading order ( pythia @xcite , herwig @xcite ) , and higher order tree - level ( alpgen @xcite , sherpa @xcite ) generators , both coupled with parton showers . in the first paper @xcite , the jet @xmath19 is presented in events with up to 3 jets .
both herwig and pythia show significant shape and normalization differences when compared to the data .
in contrast , both alpgen and sherpa , which combine tree - level calculations with parton shower enhancements are in reasonable agreement with the shape of the data distributions .
the residual normalization differences suggest residual large scale uncertainties .
however , for a specific choice of scale , the normalization of either generator can be made to agree with the data .
distribution for @xmath1 + jet events .
the cited article @xcite has similar distributions for up to 3 jets.,width=302 ] a second d @xmath1 + jets analysis @xcite uses the muon decay channel and investigates the angular correlations of the @xmath1 boson and the leading jet . in this analysis ,
the distributions @xmath36 , @xmath37 , and @xmath38 are presented .
figure [ figure_d0_z_jets_2 ] shows the @xmath36 result .
the summary conclusions of the paper is that alpgen and sherpa are superior generators , although there remain scale dependences in the theory that it is hoped that tuning to this data will reduce .
distribution between the @xmath1 boson and the leading jet .
the cited article @xcite has similar distributions for other angular variables.,width=302 ]
with the turn - on of the lhc , it is naturally expected that new limits will be set for the higher jet @xmath19 and dijet mass scales . however , in order to be competitive they will need to achieve systematic uncertainties . in order to achieve the uncertainties that d quotes here required several years of study of their calorimetry .
it is expected that an extended period of time will be needed by both cms and atlas .
early qcd results from the lhc will be the charged particle spectra within jets and jet measurements that are relatively insensitive to jet energy corrections , for instance the @xmath39 distribution between the two leading jets .
the study of qcd has now come of age .
very precise measurements are available , as are calculations that can be as high as nnlo in precision .
the magnitude of the high-@xmath40 gluon contribution to the pdfs remains a question of some concern and it is hoped that the precise measurements of the tevatron will reduce the uncertainty associated with this phenomenon .
the study of @xmath41 + jets events is crucial to reduce the background systematic uncertainties for higgs boson searches , as well as other new phenomenon .
the era of the lhc is upon us .
there is considerable excitement among the field as to what lessons the lhc data will provide . within two years of turn on
, it is quite possible that the extended energy range will have revealed new information in our study of the fundamental building blocks of the universe .
j. campbell and r.k .
ellis , phys .
d * 65 * , 113007 ( 2002 ) .
t. sjstrand _
et al . _ ,
jhep * 0605 * , 026 ( 2006 )
. g. corcella _
et al . _ ,
jhep * 0101 * , 010 ( 2001 )
et al . _ ,
jhep * 0307 * , 001 ( 2003 )
. t. gleisberg _
_ , jhep * 0402 * , 056 ( 2004 ) . | the study of the inelastic scattering of hadrons has progressed in the last decade . with the availability of high - statistics data sets from hera and the tevatron ,
our understanding of high energy and high jet multiplicity events has become rather precise . in this proceedings ,
i present an overview of recent jet - only results , as well as measurements of events which combine both jets and a @xmath0 or @xmath1 boson . |
recently , a lot of interest has grown around the possibility of applying string inspired techniques to the non - perturbative regime of qcd .
the starting point is the ads / cft correspondence @xcite , a conjectured duality between a maximally supersymmetric strongly coupled conformal field theory and the supergravity limit of type iib string theory , which involves theories different from qcd .
further developments @xcite have tried to apply the correspondence to qcd , induced by the evidence of the existence of a window of energy in which qcd shows an approximate conformal behaviour @xcite .
these developments have taken different directions .
the framework through which i move here is the so - called soft wall model of ads / qcd @xcite , a phenomenological model originally built to holographically describe chiral symmetry breaking and then adapted to several strong interaction processes . for a list of other approaches the reader can refer to @xcite . in the following ,
i discuss the scalar glueball sector and how the spectrum and the two - point correlation function are represented in the soft wall model .
then , i comment on the results , comparing them with current phenomenology and lattice data .
the considered model is defined in a @xmath0 curved space ( the bulk ) with metric : @xmath1 with @xmath2 ; @xmath3 is the ads curvature radius , and the coordinate @xmath4 runs in the range @xmath5 .
qcd is supposed to live on the boundary @xmath6 , where the element @xmath7 describes a flat minkowski space .
in addition to the ads metric , the model is characterized by the presence of a background dilaton field : @xmath8 exponentially coupled to the fields , whose functional form is chosen in such a way to have linear regge trajectories for light vector mesons @xcite ; @xmath9 is a dimensionful parameter setting the scale of qcd quantities and it is of @xmath10 .
it is the responsible of the breaking of conformal symmetry and it is fixed by the experimental slope of the rho mesons trajectory .
@xmath11 glueballs can be described in qcd by the dimension four operator @xmath12 $ ] . in the five dimensional theory its dual field is a massless scalar @xmath13 @xcite , whose action is given by : @xmath14 where @xmath15 is a parameter introduced to give the correct dimension to the action .
the ads / cft dictionary states that this action is equivalent to the qcd partition function , in which the source of @xmath16 is the boundary value @xmath17 of the field @xmath13 .
the following relation can be written : @xmath18 where the function @xmath19 is called bulk - to - boundary propagator , since it links the fields in the bulk with the sources on the boundary .
it is possible to obtain qcd correlation functions functionally deriving the action ( [ action ] ) with respect to @xmath17 .
the two - point function obtained in this way is , in the limit @xmath20 @xcite : @xmath21+\non\\ & & + q^2\left[-\fr{\n^2}{2}+\fr{c^2}{4}\left(1 - 4\g_e+2\ln4 - 2\ln(q^2/\n^2)\right)\right]+\\ & & -\fr{5c^4}{6}+\fr{2c^6}{3q^2}+{\cal o}\left(\fr{1}{q^4}\right)\biggr\}\;\;,\non\end{aligned}\ ] ] to be compared with the qcd result @xcite : @xmath22 matching ( [ piads ] ) with ( [ piqcd ] ) ( @xmath23 ) one gets the fully analytic form of the correlator . by casting it in the form : @xmath24 it is possible to find the poles @xmath25 and the related residues @xmath26 , corresponding to the mass spectrum and the decay constants of the scalar glueballs . the results for the lowest lying state are : another point is that in the large @xmath31 expansion there is a dimension two condensate , absent in qcd since there are no ways to construct scalar local gauge invariant quantities with that dimension @xcite .
0 j. m. maldacena , adv .
theor . math .
* 2 * , ( 1998 ) 231 ; e. witten , adv .
* 2 * , ( 1998 ) 253 ; s. s. gubser _ et al . _ ,
b * 428 * , ( 1998 ) 105 ; e. witten , adv .
* 2 * , ( 1998 ) 505 .
i. r. klebanov and e. witten , nucl .
b * 556 * ( 1999 ) 89 ; j. polchinski and m. j. strassler , phys .
* 88 * ( 2002 ) 031601 .
o. andreev , phys .
d * 73 * , ( 2006 ) 107901 ; a. karch _
et al . _ ,
d * 74 * , ( 2006 ) 015005 .
u. gursoy and e. kiritsis , jhep * 0802 * ( 2008 ) 032 ; u. gursoy _ et al .
_ , jhep * 0802 * ( 2008 ) 019 ; j. erdmenger _
et al . _ ,
j. a * 35 * ( 2008 ) 81 .
v. a. novikov _ et al .
_ , nucl .
b * 165 * ( 1980 ) 67 ; nucl .
b * 191 * ( 1981 ) 301 ; p. pascual and r. tarrach , phys .
b * 113 * ( 1982 ) 495 ; c. a. dominguez and n. paver , z. phys .
c * 31 * ( 1986 ) 591 ; s. narison , nucl .
b * 509 * ( 1998 ) 312 .
c. j. morningstar and m. j. peardon , phys .
d * 60 * ( 1999 ) 034509 . | i describe scalar glueballs in the soft wall model of holographic qcd ( ads / qcd ) .
[ 1999/12/01 v1.4c il nuovo cimento ] |
approximately 1 - 3% of single white dwarfs with cooling ages less than 0.5 gyr possess an infrared excess resulting from a circumstellar disk ( farihi et al .
these same stars also have distinctively high atmospheric calcium abundances ( kilic et al . 2006 , jura et al .
2007a ) even though photospheric metals normally are absent in white dwarfs in this temperature range ( zuckerman et al .
the disks likely are caused by the tidal disruption of asteroids , and there is subsequent metal accretion onto the white dwarfs ( see jura 2008 ) .
polluted white dwarfs and their disks are powerful tools to investigate extrasolar asteroids the building blocks of extrasolar rocky planets .
there are 14 single white dwarfs currently known to have a definite or highly - likely continuum infrared excess ( zuckerman & becklin 1987 , becklin et al .
2005 , kilic et al .
2005 , 2006 , jura et al
. 2007a , kilic & redfield 2007 , von hippel et al .
2007 , farihi et al .
2008a , b , brinkworth et al .
2008 , melis et al .
2008 , in preparation ) .
previously , spectra with the infrared spectrograph ( irs , houck et al .
2004 ) on the _ spitzer space telescope _
( werner et al . 2004 ) have been reported for two stars : g29 - 38 ( reach et al .
2005 , reach et al .
2008 ) and gd 362 ( jura et al .
2007b ) . both stars exhibit a strong 10 @xmath0 m emission feature whose red wing can be modeled as arising from olivine - like glasses . here , we report irs results for 6 additional white dwarfs with an infrared excess .
irs spectra would be desirable for all white dwarfs with a definite or highly - likely continuum infrared excess .
we observed the first 6 white dwarfs listed in table 1 .
an irs spectrum was obtained for ltt 8452 by another group but never published ; for completeness , we report the results here .
we did not target g166 - 58 because a nearby bright background galaxy precludes useful observations at @xmath1 @xmath2 8 @xmath0 m .
sdss 1228 + 1040 ( gaensicke et al . 2006 , brinkworth et al .
2008 ) , pg 1457@xmath3086 ( farihi et al .
2008b ) and ton 345 ( melis et al .
2008 , in preparation ) , were not known to have an infrared excess at the time the present program was implemented .
observations with irs were executed during _ spitzer _
cycle 4 , between 2007 july and 2008 february .
spectroscopy was performed in staring mode using modules sl1 and sl2 which essentially cover the 515 @xmath0 m region with spectral resolution between 64 and 128 .
the respective ramp times for these targets in each module are listed in table 1 and were constrained by the expected sky backgrounds in the region of each target .
longer ramp times for these faint sources were prohibited by the predicted backgrounds lest the data become problematic due to saturation in the peak - up sections of the array ( _ spitzer _ science center 2007 ) .
additionally , irs observations of the white dwarf ltt 8452 , performed during cycle 2 in 2006 october , were extracted from the _ spitzer _ archive for analysis .
the spectral data were processed with the irs calibration pipeline , versions 15 ( ltt 8452 ) , 16 ( ec 11507@xmath31519 and gd 56 ) , and 17 ( gd 16 , gd 40 , gd 133 , and pg 1015 + 161 ) . the combined , sky - subtracted frames were manually corrected for bad pixels using the irsclean package , then processed with spice to perform spectral extraction . the optimal extract function was used for these relatively low signal - to - noise ( s / n ) data ; first with the default aperture function ( 8 pixels at 12 @xmath4 m ) to assess the proper calibration level , then with a custom aperture function ( 4 pixels at 12 @xmath4 m ) to produce a higher s / n extraction .
data at both nod positions were averaged and the custom extraction data were scaled to the well - calibrated level of the default extraction , with all orders combined and averaged in regions of overlap .
no inter - order adjustments were made to the data .
the s / n can be estimated by examining the spectra themselves and evaluating the deviation within regions where a featureless , flat continuum might be expected . in this manner ,
the s / n over the 911 @xmath4 m region was estimated by taking the mean flux in that region and dividing by the standard deviation in the 6 - 8 @xmath4 m region .
the spectra shown in figures 16 have s / n between 3.5 and 8.0 in the 911 @xmath4 m interval .
the irs dataset for ltt 8452 is somewhat problematic ; it appears to suffer from saturation effects in the peak - up portions of the array due to high background , and possibly from point source overexposure . while the latter does not produce an unrecoverable problem for irs data , the former has the potential to significantly compromise the s / n and the spectrophotometric flux calibration . in the co - added spectroscopic bcd frames , the median flux levels in the blue and red sub - arrays are just above 850 and 1050 electrons s@xmath5 , respectively , corresponding to point source flux levels of more than 0.1 jy , a value which saturates the blue sub - array in a ramp time of 14 seconds , and the red sub - array in around 30 seconds ( _ spitzer _ science center 2007 ) .
at least two bright point sources are seen in the peak - up arrays in the nod - subtracted spectroscopic frames , and it is likely the overall background level exceeded the recommended 25 mjy / sr limit for 240 sec ramp times ( consistent with the background predictions made by spot for this object ) . in any case
the spectroscopic signal is unaffected by flux bleeding from the peak - up sub - arrays ; hence one can confidently say there are no features in the spectrum of ltt 8452 which are the result of the overexposure .
however , the s / n is clearly compromised relative to expectations , and the overall calibration level does not match its irac photometry .
the spectrum of ltt 8452 has been re - scaled up by a factor of 2.5 to match these fluxes .
no detectable signal was present in the irs observations of pg 1015 + 161 .
it is likely the source was too faint for the short ramp times required by the relatively high background region of the sky .
the resulting spectra with both smoothed and unsmoothed data along with relevant photometry are displayed in figures 16 .
additionally , for each star , we present the profile of the interstellar silicate feature inverted from absorption to emission ( kemper et al . 2004 ) , the profile of the silicate emission in bd + 20 307 , a main - sequence debris disk star with a remarkably strong 10 @xmath0 m feature ( song et al .
2005 ) , and a model of the circumstellar disk emission described in more detail below .
according to min et al .
( 2007 ) , the interstellar silicate profile of kemper et al .
( 2004 ) is generally representative of diffuse interstellar matter . in the study by wyatt et al .
( 2007 ) of main - sequence debris disks with warm dust that emits near 10 @xmath0 m , bd + 20 307 has by far the largest fraction of stellar luminosity that is reprocessed by dust in a system where the grains can not be primordial .
thus the infrared spectrum of bd + 20 307 likely is our best current measure of the spectrum of dust derived from extrasolar rocky parent bodies .
rhee et al .
( 2008 ) have reported an infrared excess around hd 23514 , an f - type main - sequence star in the pleiades , which , like bd + 20 307 , also reprocesses more than 1% of the host star s luminosity . however , the silicate emission from hd 23514 peaks near 9 @xmath0 m , and the feature is very non - standard .
each observed white dwarf displays silicate emission that peaks near 10 @xmath0 m .
all profiles are notably better matched by the profile from bd + 20 307 than the interstellar profile , and in all cases the red wing of the feature extends to at least 12 @xmath0 m .
the white dwarf spectra are more closely matched by dust derived from rocky parent bodies than by interstellar dust .
for comparison , the smoothed data from figures 16 plus the irs data for gd 362 and g29 - 38 are shown in figure 7 .
the features are not identical ; gd 362 exhibits the greatest line - to - continuum ratio while gd 133 possesses a particularly broad feature .
we see no evidence of pah emission or any other spectroscopic features .
the continuum infrared fluxes which typically reprocess @xmath61% of a white dwarf s luminosity are fit by a flat , passive , opaque dusty disk ( jura 2003 , jura et al .
our irs observations show that typically @xmath60.1% of the stellar luminosity is reprocessed as a 10 @xmath0 m silicate emission .
while the overall disk parameters are largely determined by the continuum fluxes , we use the irs data to constrain the composition of the circumstellar dust . to exploit the irs data , we elaborate upon our previous very simple models for white dwarf dust disks . in these models ,
the inner boundary is near the location where sublimation of silicates becomes rapid while the outer boundary lies within the tidal radius of the star .
these simple models presume that the disk is vertically isothermal at every radius , as may occur in regions where the gas density is high enough for thermal conduction to be important and emission from such regions can not explain the silicate features .
there are at least two possible generalizations .
( 1 ) the top " of the disk is directly illuminated by the light from the host star and is hotter than the midplane if thermal conduction is not dominant .
disks with little gas possess a vertical temperature gradient and an emission feature can result .
( 2 ) the outer zones of the disks can be optically thin . to explain the data for gd 362 with its particularly strong silicate emission , jura et al .
( 2007b ) employed both elaborations to the simple disk in a model with three radial zones . here
, except for gd 56 described below , we match the data in figures 16 with models consisting of two radial zones .
we have found that both kinds of generalizations to the simple disk model can be made to fit the data . in order to compute models with a hot layer on top of a cold interior , we need to estimate thermal conduction by the gas as a function of location in the disk . while at least some disks contain gas ( gaensicke et al .
2006 , 2007 ) , its amount and spatial distribution are unknown . in this paper , because fewer unknown parameters are required , we only present results for calculations with an optically thin outer region . in our models ,
each radial zone is characterized by the planck - mean infrared optical depth measured vertically through the disk , @xmath7 . in the inner zone
we assume @xmath7 @xmath8 1 and that the dust is vertically isothermal . in this zone , at distance @xmath9 from the star of radius @xmath10 and effective temperature @xmath11 , the disk temperature , @xmath12 , is ( jura 2003 ) : @xmath13 the observed flux from the opaque portion of the disk depends upon its inclination angle . in the outer zone ,
we take @xmath7 @xmath14 1 . however , even though the disk is vertically transparent , since light from the incident star arrives nearly parallel to the disk surface , it is mostly absorbed .
if @xmath15 denotes the ratio of the mean opacity for absorption of optical and ultraviolet light compared to the mean opacity at infrared wavelengths where the dust emits , then at the top " of the disk where the vertical optical depth is zero , the temperature , @xmath16 , is ( jura et al .
2007b ) : @xmath17 for both the relatively large grains and the temperature range of interest here , we adopt the very rough approximation that @xmath15 = 1 .
therefore , in this zone , the vertical temperature variation is ( jura et al .
2007b ) : @xmath18 in equation ( 3 ) , @xmath19 is the mean cosine of the slant angle of the radiation incident from the star onto the disk ( jura et al .
2007b ) : @xmath20 because we take @xmath7 @xmath14 1 , emission from both sides " of the disk contributes to the observed flux which is independent of the disk inclination angle .
we list in table 2 , our adopted model parameters for each star .
as described below , the model for gd 56 is more complex and requires additional parameters .
each model requires two stellar parameters : temperature , @xmath11 and angular radius , @xmath21 where @xmath22 is the distance between the sun and the white dwarf .
there are also 5 disk parameters : inner disk radius , @xmath23 , transition radius from being optically thick to optically thin , @xmath24 , outer disk radius , @xmath25 , disk inclination angle , @xmath26 , and grain radius , @xmath27 .
we employ values of @xmath11 derived from the previous studies listed in table 2 .
we fit @xmath21 from the @xmath28 and @xmath29 photometry with the assumption that emission from the disk is unimportant at these wavelengths and by approximating the stellar spectrum as a blackbody at its effective temperature .
the disk parameters are chosen to fit the data shown in figures 16 .
we find the inner boundaries of the dust disks to be located where the grain temperature ranges between 1000 k and 1500 k according to equation ( 1 ) and therefore solids rapidly sublimate .
we take @xmath30 to fit the strength of the silicate emission and to lie within the tidal radius of the white dwarf .
following jura et al .
( 2007b ) , we assume the particles are glassy silicate spheres with an olivine stoichiometry with equal abundances of magnesium and iron ( dorschner et al .
olivines are the dominant silicates in a model of the spectrum of g29 - 38 ( reach et al .
2005 ) , while lisse et al .
( 2008 ) found that the red wing of the 10 @xmath0 m feature in the spectrum of the debris - disk binary system hd 113766 is mainly the result of olivines .
for simplicity , we assume single - size dust particles with @xmath27 between 1 @xmath0 m and 2 @xmath0 m . in order to reproduce a strong emission feature at wavelength , @xmath1
, we expect that @xmath27 @xmath31 @xmath32 .
this provides an upper limit to the particle size .
we infer a lower bound to the particle size from two arguments .
small grains are likely to be relatively hot and sublimate rapidly .
also , as shown by dorschner et al .
( 1995 ) , particles with radii larger than 1 @xmath0 m display enhanced red wings to the silicate feature as found in our data .
the model disk need not be unique ( jura et al .
2007a ) , but , here , we present results for only one set of parameters .
results are shown in figures 16 ; within the uncertainties , the models reproduce the data .
jura et al .
( 2007a ) found their flat disk models for gd 56 under - predicted the measured fluxes between 3.6 @xmath0 m and 7.9 @xmath0 m . since the infrared spectrum of gd 56 displays silicate emission similar to that found in other white dwarf disks , dust reprocessing of light from the central star rather than some other source in the system seems likely . here ,
we show that a passive , warped disk can account for the data .
armitage & pringle ( 1997 ) have discussed warped disks , and jura et al .
( 2007b ) applied their models to a portion of the disk around gd 362 .
asymmetries in the double - peaked optical emission lines seen in two disks ( gaensicke et al . 2006 , 2007 ) indicate deviations from flat systems with material in circular orbits . here
, we consider a warped disk with a three - radial - zone model .
the innermost zone of the disk is flat , opaque , vertically - isothermal , and unwarped .
there is a transition radius , @xmath33 , where the disk is also opaque and vertically - isothermal , but warped . if @xmath34 is the cosine of the angle between the warped portion the disk and the line of sight to the central star so that @xmath34 = 0 corresponds to no warp , then in this portion of the warped disk : @xmath35 we propose that the outer zone of the disk is optically thin but also warped . with the assumption that @xmath15 @xmath36 1
, the temperature variation within the outer zone of disk is modified from equation ( 3 ) to : @xmath37 in this warped disk case , only the top half of the system is illuminated even in the optically thin outer zone .
we consider a parameter regime where the warp angle is sufficiently small that the solid angle subtended by the infrared - emitting disk is not appreciably changed .
the warped portions of the disk receive more light and are warmer than if the disk was flat .
consequently , the model can account for more near - infrared emission than predicted for a flat disk .
we show in figure 3 , with the warping parameters listed in the caption , a model that reproduces the data for gd 56 .
above , we have described disk models for the infrared excess seen at six white dwarfs .
a natural explanation for such a disk is the tidal disruption of an asteroid that strayed close to the star .
an alternative model to explain circumstellar dust orbiting a white dwarf is that grains are accreted from the interstellar medium and then drift inwards under the action of poynting - robertson drag .
such a model can be tested for white dwarfs where hydrogen is the dominant light element since the dwell time of metals in the atmosphere is sufficiently short that accretion must be ongoing .
specifically , four white dwarfs in table 4 have atmospheres where hydrogen is the dominant light element , and the atmospheric dwell times for calcium are 10 days , 7 days , 9 days and 200 years , for gd 56 , gd 133 , ec 11507@xmath31519 and ltt 8452 , respectively ( koester & wilken 2006 ) . using the metal accretion rates derived by koester & wilken ( 2006 ) , jura et al .
( 2007a ) showed that the measured value of f@xmath38(24 @xmath0 m ) was lower than predicted by this interstellar accretion model by factors of 40 and 140 for gd 56 and gd 133 , respectively . extending their analysis ( equation [ 3 ] in jura et al .
2007a ) to ltt 8452 , and using the stellar distance and accretion rate given by koester & wilken ( 2006 ) , the interstellar accretion model over - predicts f@xmath38(24 @xmath0 m ) by more than a factor of 1000 .
there is no reported measurement of f@xmath38(24 @xmath0 ) for ec 11507@xmath31519 , so we do not consider this star .
when relevant data are available , the tidally - disrupted asteroid model is favored .
all eight white dwarfs with irs - measured spectra of their infrared excess exhibit a 10 @xmath0 m silicate emission feature with a red wing that extends beyond 12 @xmath0 m ( reach et al .
2005 , jura et al . 2007b , figures 16 ) .
as discussed by lisse et al .
( 2008 ) , such features are characteristic of olivines and appear in the spectra of evolved solids associated with planet formation .
all white dwarfs with an infrared excess are sufficiently hot they could excite emission from circumstellar pah s , yet none do ( this paper , farihi et al .
2008a , jura et al .
2007b , reach et al .
2005 , reach et al .
the mid - infrared spectra of 25% of herbig ae stars with @xmath11 near 10,000 k are dominated by pah emission and they do not show any silicate features ( sloan et al . 2005 ) .
at least 50% of herbig ae stars ( meeus et al .
2001 ) display at least some pah emission even if not dominant .
thus , qualitatively , the absence of pah emission in their irs spectra suggests that white dwarfs are orbited by silicate - rich , carbon - poor material .
currently , gd 40 and gd 362 , two white dwarfs with infrared silicate emission , also have a detailed measure of their atmospheric composition . the apparently low carbon fraction in the disk
is paralleled with a low carbon abundance in their atmospheres since @xmath39(c)/@xmath39(fe ) @xmath40 1.02 in gd 362 ( zuckerman et al .
2007 , table 2 ) and @xmath39(c)/@xmath39(fe ) = 0.1 in gd 40 ( wolff et al . 2002 ) . for reference , in the sun , @xmath39(c)/@xmath39(fe ) = 9.3 ( lodders 2003 ) .
the rocky material in the inner solar system is carbon deficient . in ci chondrites
the most primitive of meteorities and in earth s mantle , @xmath39(c)/@xmath39(fe ) is 0.9 and 0.009 , respectively ( lodders 2003 , mcdonough & sun 1995 )
. our new data are consistent with the view that extrasolar asteroids are carbon - deficient by a factor of 10 or more ( jura 2006 , 2008 )
. therefore the building blocks of extrasolar rocky planets have the same substantial carbon deficiency as found in the inner solar system .
we report 10 @xmath0 m silicate emission from dusty disks orbiting six externally - polluted white dwarfs .
the red wings extending to 12 @xmath0 m are characteristic of micron - sized glasses with an olivine stoichiometry , particles that likely arise from tidally - disrupted asteroids .
the strong silicate emission and the absence of pah emission support the scenario that extrasolar rocky planets are assembled from silicate - rich , carbon - poor solids .
this work has been partly supported by nasa grants to ucla and is based on observations made with the _
spitzer space telescope _ which is operated by the jet propulsion laboratory , california institute of technology , for nasa .
armitage , p. j. , & pringle , j. e. 1997 , , 488 , l47 becklin , e. e. , farihi , j. , jura , m. , song , i. , weinberger , a. j. , & zuckerman , b. 2005 , , 632 , l119 boersma , c. et al .
2008 , , 448 , 241 brinkworth , c. et al .
2008 , submitted dorschner , j. , begemann , b. , henning , th . , jager , c. , & mutschke , h. 1995 , , 300 , 503 farihi , j. , zuckerman , b. , & becklin , e. e. 2008a , , 674 , 431 farihi , j. , jura , m. , & zuckerman , b. 2008b , submitted to apj gaensicke , b. t. , marsh , t. r. , southworth , j. , & rebassa - mansergas , a. 2006 , science , 314 , 1908 gaensicke , b. t. , marsh , t. r. , & southworth , j. 2007 , , 380 , l35 houck , j. et al .
2004 , , 154 , 18 jura , m. 2003 , , 548 , l91 jura , m. 2006 , , 653 , 613 jura , m. 2008 , , 135 , 1785 jura , m. , farihi , j. , & zuckerman , b. 2007a , , 663 , 1285 jura , m. , farihi , j. , zuckerman , b. , & becklin , e. e. 2007b , , 133 , 1927 kemper , f. , vriend , w. j. , & tielens , a. g. g. m. 2004 , , 609 , 826 kilic , m. , & redfield , s. 2007 , , 660 , 641 kilic , m. , von hippel , t. , leggett , s. k. , & winget , d. e. 2005 , , 632 , l115 kilic , m. , von hippel , t. , leggett , s. k. , & winget , d. e. 2006 , , 646 , 474 koester , d. , napiwotzki , r. , voss , b. , homeier , d. , & reimers , d. 2005 , , 439 , 317 koester , d. , & wilken , d. 2006 , , 453 , 1051 lisse , c. m. , chen , c. h. , wyatt , m. c. , & morlok , a. 2008 , , 673 , 1106 lodders , k. 2003 , , 591 , 1220 mcdonough , w. f. , & sun , s .- s .
1995 , chemical geology , 120 , 223 meeus , g. , waters , l. b. f. m. , bouwman , j. , van den ancker , m. e. , waelkens , c. , & malfait , k. 2001 , , 365 , 476 min , m. , waters , l. b. f. m. , dde koter , a. , hovenier , j. w. , keller , l. p. , & markwick - kemper , f. 2007 , , 462 , 667 reach , w. t. , kuchner , m. j. , von hippel , t. , burrows , a. , mullally , f. , kilic , m. , & winget , d. e. 2005 , , 635 , l161 reach , w. t. , lisse , c. , von hippel , t. , & mullally , f. 2008 , , in press ( astro - ph 0810.3276 ) rhee , j. h. , song , i. , & zuckerman , b. 2008 , , 675 , 777 sloan , g. c. et al .
2005 , , 632 , 956 song , i. , zuckerman , b. , weinberger , a. j. , & becklin , e. e. 2005 , nature , 436 , 363 spitzer science center .
2007 , spitzer observer s manual , version 8 ( pasadena : ssc ) von hippel , t. , kuchner , m. j. , kilic , m. , mullally , f. , & reach , w. t. 2007 , , 662 , 544 werner , m. w. _ et al .
_ 2004 , , 154 , 1 wolff , b. , koester , d. , & liebert , j. 2002 , , 385 , 995 wyatt , m. c. , smith , r. , greaves , j. s. , beichman , c. a. , bryden , g. , & lisse , c. m. 2007 , , 658 , 569 zuckerman , b. , & becklin , e. e. 1987 , nature , 330 , 138 zuckerman , b. , koester , d. , melis , c. , hansen , b. m. , & jura , m. 2007 , , 671 , 872 zuckerman , b. , koester , d. , reid , i. n. , & hunsch , m. 2003 , , 596 , 477 reach et al .
( 2005 ) modeled the infrared excess emission from g29 - 38 with an optically thin circumstellar cloud including @xmath63 @xmath41 10@xmath42 g of silicates . here , we show that this mass of optically thin silicates is a natural extension of a flat , largely opaque disk described in @xmath433 .
the maximum vertical thickness , @xmath44 , of any optically thin layer is determined by the cosine of the slant angle , @xmath19 , and the planck - mean opacity for absorption of this light , @xmath45 ( @xmath46 g@xmath5 ) .
if the space density of the grains is @xmath47 , then : @xmath48 with inner and outer radii such that @xmath25 @xmath8 @xmath23 , then using equation ( 4 ) , the total mass of optically thin dust , @xmath49 , on one side of the disk , is : @xmath50 for spherical grains of matter density @xmath51 and radius @xmath27 , we take : @xmath52 reach et al .
( 2005 ) consider models with grains with a power law in the size distribution such that @xmath53 varies as @xmath54 and the radius range is 0.01 1000 @xmath0 m .
for such grains illuminated by white dwarfs with @xmath11 near 10,000 k , a very approximate value of the mean size for absorption is @xmath27 near 1.0 @xmath0 m . in this case , with @xmath51 @xmath36 3 g @xmath55 , then @xmath45 @xmath36 2500 @xmath46 g@xmath5 for optical and ultraviolet wavelengths where the white dwarf emits most of its light . for @xmath25
@xmath36 100 @xmath10 and with @xmath10 = 8.2 @xmath41 10@xmath56 cm ( jura 2003 ) , then @xmath49 @xmath36 2 @xmath41 10@xmath42 g. reach et al .
( 2005 ) estimate that the optically thin cloud of dust contains @xmath610@xmath57 g of which 75% is amorphous carbon to account for the continuum and 25% is silicate to explain the 10 @xmath0 m feature . their inferred mass of @xmath62.5 @xmath41 10@xmath42 g of optically - thin silicates
is naturally computed for a flat , largely opaque , passive disk which then also accounts for the continuum fluxes .
reach et al .
( 2008 ) have acquired additional irs data for g29 - 38 beyond that described in reach et al .
( 2005 ) , and they describe two models that account equally well for the data . in their first model , the dust cloud is slightly opaque , while their second model includes both a optically thick disk and an optically thin outer region composed of silicates .
this second model is similar to our models described above in @xmath58 for the six stars listed in table 2 . in this appendix
, we note that the optically thin emission may reside on top " of the opaque disk rather than beyond it . in any case ,
a major difference between the completely optically thin model and the model with both an optically thin region and a flat , opaque disk and is the total amount of circumstellar mass .
an optically thin cloud has @xmath40 10@xmath57 g while an opaque disk can have orders of magnitude more mass .
the atmospheres of gd 40 and gd 362 which are helium - rich and therefore have a particularly deep convective zone , have been contaminated by @xmath610@xmath59 g of metals ( jura 2006 , zuckerman et al . 2007 ) .
therefore , it is plausible that each of these two stars possesses a massive opaque disk in addition to an optically thin dust cloud of @xmath610@xmath57 g. in this table , @xmath11 , @xmath10 and @xmath22 are the stellar effective temperature , radius and distance , respectively and are taken from available measurements . the disk inclination angle , @xmath26 , ( defined so that @xmath26 = 0@xmath60 is face - on ) , and radial locations of various zones of the disks and the grain radius , @xmath27 , are fit to the data as described in the text . + notes : ( a ) koester et al .
( 2005 ) , ( b ) wolff et al .
( 2002 ) , ( c ) koester & wilken ( 2006 ) | _ spitzer space telescope _ spectra reveal 10 @xmath0 m silicate emission from circumstellar dust orbiting six externally - polluted white dwarfs . micron - size glasses with an olivine stoichiometry can account for the distinctively broad wings that extend to 12 @xmath0 m ;
these particles likely are produced by tidal - disruption of asteroids .
the absence of infrared pah features is consistent with a scenario where extrasolar rocky planets are assembled from carbon - poor solids . |
clustering analysis ( or hca ) is an extensively studied field of unsupervised learning . very useful in dimensionality reduction problems , we will study ways of using this clustering method with the aim of reducing ( or removing ) the need for human intervention .
this problem of human intervention stems from the fact that hca is used when we do not know the correct number of clusters in our data ( otherwise we might use , say , k - means ) .
while the ability to cluster data with an unknown number of clusters is a powerful one , we often need a researcher to interpret the results - or cutoff the algorithm - to recover a meaningful cluster number .
while our motivation stems from dna micro - array data and gene expression problems , these methods can apply to any similarly structured scenario . specifically
, we will analyze different existing automated methods for cutting off hca and propose two new ones . in section ii we will discuss background material on hca and the existing methods and in section iii we will present some technical details on these methods and introduce our own .
section 4 will contain results on simulated and actual data , and section 5 will examine data sampling procedures to improve accuracy .
hierarchical clustering , briefly , seeks to pair up data points that are most similar to one another . with the agglomerative ( or bottom - up ) approach , we begin with @xmath0 data points forming singleton clusters .
for each point , we measure the distance between it and its @xmath1 neighbors . the pair with the shortest distance between them is taken to form a new cluster .
we then look at the distance between the @xmath2 points remaining and the newly formed cluster , and again pair off the two with shortest distance ( either adding to our 2-cluster , or forming another one ) .
this process is repeated until we have a single cluster with @xmath0 points ( regardless of the absolute distance between points ) .
naturally , this is a very good dimensionality reduction algorithm .
unfortunately , it keeps going until we ve flattened our data to 1 dimension . in cases where in truth we have @xmath3 clusters ,
this is problematic .
the results of a hca are often expressed as a dendrogram , a tree - like graph that contains vital information about the distances measured in the clustering and the pairings generated .
an example of a dendrogram can be seen in figure [ fig:1 ] .
briefly , horizontal lines denote pairings , and the height of those lines represent the distance that needed to be bridged in order to cluster the points together .
that is , the smaller the height ( or jump ) of a pairing , the closer the points were to begin with .
our goal is to find a way to say , once the full tree is made ,
jumps beyond this point are not reasonable " , and we can cutoff the algorithm , keeping only those clusters generated before that point .
the problem of cutting off a dendrogram is one that researchers encounter often , but there are no reliable automated methods for doing it .
often , the gap statistic is the only proposed automated method , as in @xcite . as such
, many researchers will inspect the finished dendrogram and manually select a cutoff point , based on their own judgment .
apart from the obviously slow nature of this exercise , there is also the question of human error to consider - as well as bias . in cases where the cutoff is not easily determined , two different researchers may arrive at different conclusions as to the correct number of clusters - which could both be incorrect .
algorithmic approaches aim to eliminate this , and range from simpler methods to more complex ones .
an excellent summary of existing methods is given in @xcite , which is in fact referenced in @xcite .
the latter , more importantly , develops the gap statistic .
we will present the technical aspects in section iii , but we quickly discuss some properties here .
first , the gap statistic is one of few methods that is capable of accurately estimating single clusters ( in the case where all our data belongs to one cluster ) , a situation often undefined for other methods .
while it is rather precise overall , it requires the use of a reference distribution " , which must be chosen by the researcher .
they put forward that the uniform distribution is in fact the best choice for unimodal distributions .
a powerful result , it is still limited in other cases , and thus many researchers still take the manual approach
. however , it generally outperforms other complex methods , as such we focus on the gap statistic . on the other side of the complexity spectrum
, we have variants of the elbow method " .
the elbow method looks to explain the variance in the data as a function of the number of clusters we assign .
the more clusters we assign , the more variance we can explain .
however , the marginal gain from adding new clusters will begin to diminish - we choose this point as the number of clusters .
a variant of this method , often applied to dendrograms , looks for the largest acceleration of distance growth @xcite .
while this method is very flexible , it can not handle the single - cluster case .
we will look at both the elbow method variant and the gap statistic , as well as our own 2 methods we are presenting in this paper . while there are many other methods to compare them to
, the gap statistic is quite representative of a successful ( if more complex ) solution - and tends to outperform the other known methods .
the elbow method represents the more accepted simple approaches . in all tests in this paper
, we use an agglomerative hierarchy , with average linkage and euclidean distance measure .
the gap statistic is constructed from the within - cluster distances , and comparing their sum to the expected value under a null distribution .
specifically , as given in @xcite , we have for @xmath4 clusters @xmath5 @xmath6 - \log w_k\ ] ] where , with @xmath7 , @xmath8 that is , we are looking at the sum of the within - cluster distances @xmath9 , across all @xmath4 clusters @xmath5 .
computationally , we estimate the gap statistic and find the number of clusters to be ( as per @xcite ) @xmath10 where @xmath11 is the standard error from the estimation of @xmath12 .
as mentioned , @xcite considers both a uniform distribution approach and a principal component construction . in many cases , the uniform distribution performs better , and this is the one we will use .
this variant of the elbow method , which looks at the acceleration , is seen in @xcite . a straightforward method
, we simply look at the acceleration in jump sizes .
so given the set of distances from our clustering @xmath13 , the acceleration can be written as @xmath14 we choose our number of clusters as the jump with the highest acceleration , giving us @xmath15 } \left\ { d_i-2d_{i-1}-d_{i-2 } \right\}.\ ] ] while very simple and very fast , this method will never find the endpoints , ie , the @xmath0 singleton clusters and the single @xmath0-element cluster cases .
the first method we propose is based on the empirical distribution of jump sizes .
specifically , we look to the mode of the distribution @xmath16 , denoted @xmath17 , adjusted by the standard deviation ( @xmath18 ) .
our motivation is that the most common jump size likely does not represent a good cutoff point , and we should look at a higher jump threshold . as such
, we take the number of clusters to be @xmath19 where @xmath20 is a parameter that can be tuned .
naturally , tuning @xmath20 would require human intervention or a training data set .
as such , we set it to @xmath21 somewhat arbitrarily .
our second method is even simpler , but is surprisingly absent from the literature .
inspired by the elbow method , we look at the maximum jump difference - as opposed to acceleration . our number of clusters
is then given by @xmath22}\left\ { d_i - d_{i-1 } \right\}.\ ] ] this method shares the elbow method s drawback that it can not solve the single cluster case ( though it can handle singleton clusters ) , but we thought it prudent to examine as the literature seemed to focus on acceleration and not velocity .
we present results of simulations on several numbers of true clusters , drawn from a @xmath23-dimensional normal distribution .
each cluster is comprised of @xmath24 samples .
we are most interested in tracking the success rate and the _ error size given an incorrect estimate_. that is , how often can we correctly estimate the number of clusters @xmath25 and when we ca nt , by how much are we off ?
formally , this is given by @xmath26 and @xmath27 we chose this measure of the error to avoid under - estimating error size ( which would happen in a method that is often correct ) .
the data used was drawn from a standard normal distribution , with cluster centers at @xmath28 , shown in figure [ fig:2 ] . in the case of @xmath29 cluster , the first is taken , for @xmath23 clusters , the first two , and so on .
we present the results of the methods on @xmath30 simulations below in tables i - iii , with the best results in bold . .average cluster numbers over @xmath30 runs . [ cols="<,>,>,>,>",options="header " , ] while the maximum difference method seems robust in the face of different sampling , it seems that this exercise has revealed some instability in the mode method , which has reverted back to a lackluster performance . to a much lesser extent ,
the elbow method is having some more trouble as well .
it seems more likely that the choice of sampling parameters could be the cause of the clustering in the biobase data in figure [ fig:6 ] .
more generally , we should look into determining optimal mixing parameters @xmath31 and @xmath32 and/or their impact on these methods .
this mixing method does appear to perform better for the gene expression set than the previous choice of mixing parameters , which seems to confirm our hypothesis that there is perhaps an oversampling effect that can come into play , or something along those lines which much be explored more fully .
we have developed two new empirical methods for clustering data in a hierarchical framework . while our methods are substantially faster than the existing gap statistic , we insist that they do not handle the single - cluster case .
in other cases , our maximum difference method is at least comparable to the gap statistic and outperforms the elbow method .
in addition , the use of the data mixing procedure presented can greatly improve performance ( especially for the mode method ) , leading to the maximum difference method outperforming the other 3 .
lastly , these methods can be implemented in a few lines of code and should allow researchers to quickly utilize them at little computational cost . in the future
we hope to study the possibility of finding optimal mixing numbers @xmath31 and @xmath32 and the impact of the choice of these parameters of our results .
hopefully , they are related to the instability detected in the mode method when using @xmath33 in our mixing procedure . | we propose two new methods for estimating the number of clusters in a hierarchical clustering framework in the hopes of creating a fully automated process with no human intervention .
the methods are completely data - driven and require no input from the researcher , and as such are fully automated .
they are quite easy to implement and not computationally intensive in the least .
we analyze performance on several simulated data sets and the biobase gene expression set , comparing our methods to the established gap statistic and elbow methods and outperforming both in multi - cluster scenarios . shell : bare demo of ieeetran.cls for journals clustering , hierarchy , dendrogram , gene expression , empirical . |
about 95% of the energy contents of the universe appears to be of unknown origin , in the form of dark matter and dark energy . while there is a lot of astrophysical evidence for the existence of dark matter and dark energy , a deeper understanding of the physical nature of these main ingredients of the universe is still lacking .
clearly it is important to design new experiments on earth that could have the potential to unravel some of the unknown physics underlying dark matter and dark energy . at the particle physics level
, there are two main candidates what dark matter could be .
these are wimps ( weakly interacting massive particles ) @xcite and axions @xcite .
wimps are motivated by supersymmetry , whereas axions are motivated by the solution of the strong cp problem in qcd .
various experimental searches to detect wimps @xcite and axion - like particles @xcite on the earth are currently going on . very recently , there have been a couple of new suggestions how one could possibly detect dark matter axions in laboratory experiments on the earth @xcite .
all these proposals have in common that they are based on relatively small devices and that they suggest to look for small oscillating electric currents induced by axion flow , with a frequency given by the axion mass .
proposal 1 @xcite is based on a technique similar to nuclear magnetic resonance ( nmri ) , known from medical imaging .
proposal 2 @xcite is based on resonance effects in josephson junctions .
proposal 3 @xcite suggests to use lc circuits cooled down to mk temperatures .
further interesting proposals are based on topological magnetic insulators @xcite and atomic systems @xcite . in this paper
we present a detailed calculation describing the physics of proposal 2 , starting from the field equations of axion electrodynamics in a josephson environment .
in contrast to axions in vacuum , in a josephson junction the axion has the possibility to induce electric supercurrents , rather than just ordinary currents .
our main result presented in this paper is that , besides the trivial solution where the axion passes through the josephson junction without interaction , there is a nontrivial solution to the axion field equations due to these supercurrents .
we show that the nontrivial solution implies the existence of a huge ( formal ) axion - flow generated magnetic field in a tiny surface area of the weak - link region of the junction , which makes incoming axions decay into microwave photons .
the axion flow from the galactic halo through the junction then leads to a small measurable excess current of cooper pairs , for which we will derive a concrete formula .
the experimental consequence of this are shapiro steps @xcite generated by axion flow , which are small but observable provided certain conditions on the design of the josephson junction are satisfied .
we will derive these conditions explicitly .
an experiment by hoffmann et al . based on s
/ n / s josephson junctions @xcite , discussed in detail in @xcite , provided evidence for an axion mass of 110 @xmath1 and an axionic dark matter density of about 0.05 gev/@xmath2 if interpreted in this way . here
we will discuss the results of four different experiments @xcite . in all four cases small shapiro step - like anomalies
have been observed that , if interpreted within our theory , point towards an axion mass of @xmath3ev .
the predicted axion mass value has profound cosmological implications .
if this value is confirmed by further experiments , it means that the peccei - quinn symmetry breaking took place _ after _ inflation @xcite .
employing the recent results of @xcite our result implies that the fractional contribution @xmath4 to the cosmic axion density from decays of axionic strings and walls is @xmath5 .
this paper is organized as follows : in section 2 we write down the axion field equations in a josephson junction .
the nontrivial solution , where the axion - induced electric current manifests itself as a supercurrent within the junction , is discussed in section 3 .
the physical interpretation of this solution is further worked out in section 4 . in section 5
we present a short calculation how s / n / s josephson junctions should be designed in order to serve as optimum axion detectors .
section 6 discusses some experimental candidate signals seen in various josephson experiments that could possibly be associated with the nontrivial solution of section 3 .
section 7 compares our mass estimate from josephson resonances with cosmological and astrophysical bounds on the axion mass .
finally , our concluding remarks are given in section 8 .
let us consider the classical field equations of axion electrodynamics @xcite in a josephson junction ( jj ) @xcite .
@xmath6 denotes the misalignement angle of the axion field @xmath7 , @xmath8 the electromagnetic phase difference in the jj . in si units
one has @xmath9 here @xmath10 denotes the axion mass , @xmath11 is the axion coupling constant , @xmath12 is a tiny damping constant , @xmath13 is the electric field , @xmath14 is the magnetic field , @xmath15 is a coupling constant of order 1 ( @xmath16 for ksvz axions @xcite , @xmath17 for dfsz axions @xcite ) , @xmath18 is the fine structure constant , @xmath19 is the critical current of the junction , @xmath20 an external driving current , @xmath21 is a small axion - induced electric current in the junction , @xmath22 is the normal resistance of the junction , and @xmath23 its capacity . as usual ,
@xmath24 and @xmath25 denote electric current and charge densities .
the expected mass of the qcd dark matter axion is in the region @xmath26ev to @xmath27ev due to astrophysical and cosmological constraints ; the corresponding compton wave length is much larger than the typical size of a jj .
thus we may neglect spatial gradient terms @xmath28 in the above equations and consider the axion field as being approximately spatially homogeneous .
the most important axion contribution for detection purposes comes from the last term in eq .
( [ 2 ] ) : in a magnetic field @xmath14 , temporal changes @xmath29 of the axion angle imply an axion - induced electric current density @xmath30 given by @xmath31 note that this current is in the direction of the magnetic field @xmath14 , and not orthogonal to it , as in ordinary electrodynamics .
dark matter axions correspond to an oscillating solution of eq .
( [ 1 ] ) with @xmath32 and @xmath12 negligible , given by @xmath33 the frequency @xmath34 is given by the axion mass .
the dark matter energy density due to axions , @xmath35 , is related to the amplitude @xmath36 of the oscillations by @xmath37 this can be used to eliminate @xmath36 as @xmath38 @xmath36 is very small : the astrophysical estimates of dark matter density in the halo @xcite give something of the order @xmath39 . now consider as a suitable axion detector a driven jj in the voltage stage which contains a constant magnetic field @xmath14 near the surface of the weak link ( wl ) region that points in the direction of the bias current @xmath20 of the junction . at the moment
we do not discuss the origin of this magnetic field , it can be an external magnetic field , though later we will see that as a consequence of the field equations axions can formally self - induce such a magnetic field at the surface of wl .
we denote the distance between the two superconducting electrodes of the junction by @xmath40 , the width of the superconductors by @xmath41 and their height by @xmath42 , so that the volume of the weak link region is @xmath43 .
the volume of the region where the magnetic field is present is @xmath44 , we assume @xmath45 , i.e. the magnetic field is only present near the surface of wl .
let us consider axions from the galactic halo that enter wl transversally with velocity @xmath46 through the plane spanned up by @xmath41 and @xmath40 .
the dark matter oscillations ( [ 6 ] ) yield via eq .
( [ 5 ] ) an axion - induced current @xmath47 which couples into the jj via eq . ( [ 4 ] ) .
@xmath47 is given by @xmath48 here @xmath49 is the small surface area through which the magnetic field penetrates .
@xmath50 is an oscillating current produced by entering dark matter axions coming from outside wl .
what happens inside wl will be discussed in the next section .
the main difference between axions in vacuum and axions in a jj is that in a jj there is the possibility that electric currents are supercurrents , i.e. they manifest themselves in form of cooper pairs that tunnel the weak - link region .
this possibility is also open to electric currents induced by dark matter axions in a magnetic field .
the main assumption in the following is that the oscillating current ( [ 8 ] ) manifests itself as a supercurrent in the jj .
this means we can write for @xmath51 @xmath52 where @xmath53 can be regarded as an axion - generated critical current in the junction , and the phase @xmath54 grows linearly inside wl .
note that @xmath55 has different behavior for @xmath56 wl and @xmath57 wl .
we will later see what the physical interpretation of this is ( in fact inside wl the axion will decay immediately so that the phase @xmath58 will get the meaning of an ordinary electromagnetic phase ) . at the moment
we just do the maths , i.e. regard @xmath58 as a solution of the equations and come to the physical interpretation later .
equality between eq .
( [ 8 ] ) and ( [ 9 ] ) implies that inside wl the variable @xmath58 evolves in a different way than outside wl , namely , we have inside wl @xmath59 and @xmath60 apparently , coincidence of the electric currents ( [ 9 ] ) , ( [ 8 ] ) produced in a small vicinity inside and outside wl requires that the angle variable @xmath54 switches from harmonic oscillations to linearly increasing behavior modulo @xmath61 when entering wl , still with the same frequency @xmath62 . note that the phase difference @xmath8 of the jj anyway increases linearly in time , i.e. @xmath63 , where @xmath64 is the josephson frequency , since the jj is in a driven voltage stage @xcite .
this means that if @xmath65 the linear change of @xmath8 agrees with that of @xmath58 , which is physically absolutely reasonable .
the linear change @xmath66 from eq .
( [ here ] ) still has to respect the axion field equations .
it couples back into the system via eq .
( [ 1 ] ) .
suppose the effective damping constant @xmath12 in eq .
( [ 1 ] ) ( at the surface of the junction ) satisfies @xmath67 this is not a very restrictive condition , since for axions with a mass of @xmath68 one has the very small value @xmath69 . if this condition is satisfied then the first , third and fourth term on the left - hand side of eq .
( [ 1 ] ) can be neglected , and eq .
( [ 1 ] ) reduces to @xmath70 ( if condition ( [ cond ] ) is not satisfied then there will be an additional oscillating component to the @xmath71-field , otherwise the argumentation is very similar . )
the linear increase @xmath72 inside wl due to eq .
( [ here ] ) thus formally induces an axion - generated magnetic field @xmath14 via eq .
( [ there ] ) .
using @xmath73 , where @xmath74 is the external voltage applied to the jj , as well as using the resonance condition @xmath75 between axion mass and josephson frequency @xmath76 ( which can be achieved by suitably choosing @xmath74 ) we get from eq .
( [ there ] ) @xmath77 we have thus found a nontrivial solution of the field equations ( [ 1])([4 ] ) where there is an axion - generated magnetic field ( in the direction of the bias current ) given by ( [ 13 ] ) due to the existence of supercurrents .
depending on what is assumed for @xmath12 , this can be a huge magnetic field ( see @xcite for some numerical examples ) , but it is penetrating only through a tiny surface area @xmath49 so that the flux is reasonably sized . to properly describe axion electrodynamics in a jj , we also need to take into account the probability of axion decay in a strong magnetic field . from the primakov effect one
has for the decay probability of axions in a magnetic field of strength @xmath71
@xcite @xmath78 here @xmath42 is the length of the detector , @xmath79 is the axion - photon momentum transfer , and @xmath80 . in particular
, for the length scale @xmath81 within which the axion decays with probability @xmath82 one has ( for @xmath83 ) @xmath84 clearly , the axion - generated @xmath14 field is only present in the area where the axion still exists and has not yet decayed .
this suggests to use for the area @xmath85 in equation ( [ 11 ] ) the value @xmath49 , where @xmath81 is given by ( [ 15 ] ) .
putting eq .
( [ 13 ] ) , ( [ 15 ] ) and ( [ 17 ] ) into eq .
( [ 11 ] ) a remarkable simplification takes place and one finally ends up with the simple formula @xmath86 note that physically unmeasurable quantities like the formal huge magnetic field concentrated in a tiny surface area as well as the unknown constant @xmath12 have all dropped out , and the critical current @xmath53 is basically determined by the dark matter axion velocity @xmath46 relative to the junction and the axionic dark matter density near the earth @xmath35 , as well as the fundamental constants @xmath87 , @xmath18 and @xmath88 ( the charge of cooper pairs ) .
a linearly increasing phase in ordinary jj physics means tunneling cooper pairs and emission of josephson radiation @xcite .
as we have seen , if the axion is still assumed to be present in wl , then a linearly increasing phase @xmath89 implies the existence of a huge magnetic field given by eq .
( [ 13 ] ) . alone for energetic reasons , such a huge field can not be present in the junction .
the only sensible conclusion is that the axion immediately decays when entering the junction , it tunnels the junction @xcite
. if the axion is still present in wl , then the huge magnetic field induced by ( [ 13 ] ) will make it decay immediately .
if the axion is not present in wl , then there is no huge magnetic field , and hence no energetic problem with that field . during the tunneling process
the phase @xmath54 just gets the meaning of an ordinary electromagnetic phase in the junction , which obeys standard type of squid physics @xcite .
the tunneling axion still produces a measurable effect due to an additional contribution to the critical current given by ( [ ic1 ] ) . in calculating this critical current , all singular ( unmeasurable ) quantities such as the huge formal magnetic field
have dropped out .
we used the formal magnetic field mainly as a mathematical tool to calculate the axion - generated effects in wl from the field equations . in experiments
one should look for a small peak - like structure of the differential conductance at a particular voltage @xmath74 , produced by the increase in the critical current .
this voltage satisfies @xmath90 . putting into eq .
( [ ic1 ] ) typical numbers for the estimated density @xmath35 and velocity @xmath46 of dark matter axions in the galactic halo relative to the earth one sees that the predicted contribution of axions to the critical current is small but perfectly measurable , of the order of @xmath91a .
our physical interpretation of the linearly increasing phase @xmath54 is that the axions passing through wl decay via josephson radiation , triggering at the same time additional cooper pairs to flow .
these additional cooper pairs yield a small but measurable effect if @xmath65 . in a recent preprint @xcite it was proposed that axions can have different interactions with cooper pairs in a superconductor than in the vacuum case , possibly stronger ones , since electron number is not conserved .
this is in line with the ideas presented here : in our consideration presented here and in @xcite , incoming axions decay into microwave photons and trigger at the same time the process of cooper pairs forming out of ordinary electrons , thus producing additional supercurrents in the voltage - biased junction .
the ( maximum ) additional critical current @xmath53 as allowed by axion electrodynamics is given by ( [ ic1 ] ) .
this critical current @xmath53 is derived from the axion field equation ( [ 2 ] ) and is a current of ordinary cooper pairs through an ordinary electromagnetic weak link ; it is not related to the existence or non - existence of a possible weak link in peccei - quinn symmetry as discussed in @xcite . in topological superconductors
@xcite , superconducting phases couple to electromagnetic fields with an axionic topological coupling , and chiral vortex lines can be josephson junctions .
the corresponding field theory was developed in @xcite .
axions can also induce nonlinear modulations of the electromagnetic field in topological insulators @xcite .
generally it is to be expected that axionic dark matter is much more likely to leave a detectable signal in complex systems of condensed matter physics than in vacuum .
coming back to the simple case of axions passing through wl of a jj , as with any critical current , @xmath53 represents the _ maximum _ electric supercurrent one can expect to see due to axionic dark matter passing through the junction , an idealized situation , as allowed by the classical field equations . in practice
, the current will often be smaller , depending on the experimental situation and the geometry of the josephson junction used . to work this out further ,
let us denote by @xmath92 the surface area of wl perpendicular to axion flow , which for orthogonally entering axions is given by @xmath93 .
so far we did mainly do classical axion electrodynamics .
but one also needs to take into account the particle nature of axions and cooper pairs .
following the ideas of @xcite , we may assume that each axion entering wl triggers the flow of @xmath94 cooper pairs ( fig . 2 in @xcite ) .
in an s / n / s junction @xmath94 is related to the number of andreev reflections and given by @xmath95 where @xmath96 is the gap energy of the superconductor @xcite .
the maximum observable supercurrent is constrained by the geometry of the jj and given by the number of axions hitting the wl region of surface area @xmath92 per time unit , multiplied by @xmath97 : @xmath98 for an optimum s / n / s axion detector , both formulas ( [ ic1 ] ) and ( [ ic2 ] ) should be valid , i.e. @xmath99 . by equating them we obtain @xmath100 the right - hand side is just determined by astrophysical dark matter properties , whereas the left - hand side yields a relation for the detecting jj experiment . for the aluminium junction used by hoffmann et al .
@xcite , one can readily check that condition ( [ dn ] ) is satisfied .
this experiment thus provides an optimum axion detector .
if one wants to use other s / n / s josephson junction , say with a higher gap energy @xmath96 , then naturally @xmath40 must be chosen smaller .
for the dark matter parameters advocated in @xcite , one obtains a characteristic length scale of @xmath101 .
let us now discuss the experimental consequences of our theory . as previously discussed in @xcite , the decaying axions produce photons and these produce small axion - induced shapiro steps , which are measurable for junctions with sufficiently large @xmath92 .
the main shapiro step occurs at a voltage given by @xmath102 , other integer multiples of @xmath103 may also occur if the axion - induced cooper pair flow intensity is high .
the typical step size should be given by eq .
( [ ic2 ] ) .
first , let us discuss hoffmann et al.s experiment @xcite , based on al - cu - al s / n / s junctions .
their measurement of a shapiro step - like feature of unknown origin at @xmath104ev in @xcite was used in @xcite to estimate the axion mass as @xmath105 and the axionic dark matter density near the earth as being @xmath106 , assuming orthogonal axion flow .
the velocity @xmath46 of axions traveling through the jj was assumed to be given by the value @xmath107 ( the velocity of the earth relative to the galactic halo ) .
let us now look at other experiments as well . in the experiment of golikova et al .
@xcite , based on al-(cu / fe)-al microbridges , a double - peak peculiarity of the measured differential resistance is observed ( fig . 4 in @xcite ) , with one rather constant peak occurring at @xmath108v , whereas the other peak position near 75@xmath26v is dependent on the length of the sample and the applied magnetic field .
a possible interpretation would be to interpret the first ( universal ) peak as coming from axions and the second peak as being due to a minigap produced by the proximity effect .
further measurements are needed to check this .
@xcite use w - au - w s / n / s junctions and report the observation of a large number of fractional shapiro steps without externally applied microwave radiation .
most shapiro steps occur in the temperature region 2.8 - 3.2k , and for this temperature region the strongest steps occur at @xmath109v ( see fig . 2 in @xcite ) , which is again the axion voltage @xmath103 . since the wolfram superconductors used by he et al .
have a gap energy @xmath96 that is larger by a factor 5 as compared to the aluminium superconductors used in @xcite and @xcite , the number of andreev reflections @xmath94 given in eq .
( [ n ] ) is larger , and hence a higher intensity of cooper pair flow is induced by the incoming axions .
this may be the reason that a larger number of observable shapiro steps is excited in this experiment .
in addition , the minigap structure created by the proximity effect may create further shapiro steps in this junction . finally let us discuss another experiment performed by bae et al.@xcite , which is quite different from the previous ones , in the sense that a high-@xmath110 superconductor is used , and that there is also some external forcing with microwaves with a given frequency @xmath111 .
bae et al .
@xcite investigated the occurence of shapiro steps when irradiating a micron - sized sample of bi-2212 , a high-@xmath110 crystal , which contains a stack of about @xmath112 intrinsic tunnel josephson junctions .
the superconducting layers of this crystal are separated by @xmath113 nm from each other .
the sample used by bae et al . had width @xmath114 m , hence in total the area of the wl region is given by @xmath115 .
this effective area is large enough to produce measurable axion - induced currents via eq .
( [ ic2 ] ) . from ( [ ic2 ] )
one obtains the prediction @xmath116 @xmath117 if @xmath118 and @xmath119 gev/@xmath2 is used .
bae et al .
@xcite irradiated their probe by external microwave radiation of frequency @xmath120 5ghz , 13 ghz , 18 ghz , 23ghz , and 26 ghz , respectively .
note that 26 ghz @xmath121 .
for all values of @xmath111 they observed well - pronounced integer shapiro steps at voltages @xmath122 given by @xmath123 , as expected from the rsj model @xcite , with @xmath124 integer .
however , two unexplained peculiarities occured ( see fig . 1 , data from @xcite ) : \1 . while for @xmath125 ghz only the usual integer shapiro steps were observed , at the frequency @xmath126 ghz also one additional fractional shapiro step is seen , formally with @xmath127 ( see fig .
this unusual step occurs at a voltage @xmath128v , and the step size is ( for increasing voltage ) 16.4 na .
while in principle fractional shapiro steps are possible due to non - sinusoidal contributions in the current - phase relation , it is very unusual to have only one such step ( there is none at e.g. @xmath129 and @xmath130 , and also none for the other values of @xmath111 ) .
our physical interpretation is that this unusual step is due to axion flow through the junction and that it is stabilized due to the commensurate ratio @xmath131 .
the observed step size 16.4na agrees with what is expected from eq .
( [ ic2 ] ) . \2 .
while for @xmath125 ghz basically all low-@xmath124 integer shapiro steps are observed , for @xmath132 ghz only odd-@xmath124 shapiro steps are seen ( @xmath133 ) whereas the even-@xmath124 shapiro steps @xmath134 are suppressed , so that the voltage difference between neighbored steps is @xmath128v rather than the expected 27@xmath26v ( see fig .
due to these double - voltage steps the pattern looks much more similar to a pattern generated by a ( phase - shifted ) frequency of 26 ghz rather than one of 13 ghz .
our physical interpretation is that at even @xmath124 the axion - generated shapiro steps fall onto the ordinary ones generated by the frequency @xmath111 and can compensate them , provided they have the same magnitude but opposite sign .
the observed zero - crossing step sizes for @xmath135 are indeed 16.4 na , which would allow for such a destructive interference . in summary ,
all four experiments mentioned contain peculiar josephson resonance effects associated with the voltage @xmath136v , pointing towards an axion mass of @xmath137ev , where we base our error estimate on the data of @xcite and @xcite .
additional experimental tests are of course still needed .
a typical axion shapiro step is predicted to exhibit small daily and yearly periodic oscillations in intensity , similar as in searches for wimps @xcite .
the daily oscillations are expected to come from the fact that the galactic axion flow relative to the earth is directed , the earth rotates but axions produce the strongest signal if they enter the junction transversally .
we also emphasize that it is clearly important to extend the search range of other axion search experiments , which are _ not _ based on jjs , to the mass region suggested by jjs , @xmath138ev .
a recent experimental proposal in this direction is @xcite .
let us finally discuss the cosmological consequences of an axion mass value of 0.11mev that the experiments discussed in the previous section seem to favor .
this value implies that the peccei - quinn symmetry was broken _
after _ the end of inflation @xcite , at least in the simplest models of cosmological axion production . based on the results of @xcite ( assuming that axions make up all dark matter ) , the mass value @xmath139ev translates into an axion coupling constant of @xmath140gev , a freeze - out temperature of @xmath141 mev , and for the fractional contribution @xmath4 to the cosmic axion density from decays of axionic strings and walls we obtain from the formula @xmath142 derived in @xcite the prediction @xmath5 . generally , the rather high energy scale of inflation that seems to be indicated by the recent bicep2 results @xcite puts a lower bound on the qcd axion mass of about 70@xmath26ev @xcite .
if production of axions from axionic strings is taken into account as well @xcite , then this lower bound still becomes sharper : shellard et al .
derive in @xcite for the axion coupling constant @xmath143 gev , which is equivalent to @xmath144ev . on the other hand
, the axion mass can not be too big because otherwise one would violate various astrophysical observational constraints .
for example , the sn1987a supernovae data imply a lower bound on @xmath11 given approximately by @xmath145 gev @xcite , equivalent to @xmath146mev . in @xcite
an even stronger upper bound is derived if quantum fluctuations during inflation are taken into account , @xmath147mev .
a recent analysis of galactic rotation curves @xcite based on an axionic bose - einstein condensate as dark matter is consistent with this range of an axion mass between 0.1 and 1mev , as well as with estimates of the axionic dark matter density in the halo as calculated in @xcite .
overall , it appears that the mass value @xmath148mev that is suggested by the josephson resonances fits very well into the mass range expected from cosmological and astrophysical considerations after bicep2 .
in this paper we have presented a detailed derivation why axions can generate small measurable electric currents in josephson junctions .
we started from the field equations of axion electrodynamics , plus the assumption that the axion - induced electric current can manifest itself as a supercurrent in a josephson junction .
we found a nontrivial solution of the axion field equations in the weak - link region of the junction , for which the phase grows linearly in time .
this was interpreted in terms of axions decaying via josephson radiation and triggering at the same time additional cooper pair flow through the junction .
a huge formal magnetic field appeared in our calculations , making the axions decay if they are still present at the surface of the junction , but the final result for the axion - generated additional critical current @xmath53 as given by eq .
( [ ic1 ] ) is actually independent of the precise value of this formal @xmath14-field , as it drops out of the equations .
overall the effect of the galactic axion background is small but measurable and the decaying axions produce a small shapiro step - like feature when the axion mass resonates with the josephson frequency .
we derived concrete formulas for the additional critical current @xmath53 generated by axion flow , and derived conditions for different types of josephson junctions to act as optimum axion detectors .
the measured voltage where axion - generated shapiro steps occur can be used to estimate the axion mass @xmath10 , and their intensity can be used to estimate the axionic dark matter density @xmath35 near the earth @xcite .
we discussed peculiarities in the shapiro step patterns measured by four different experimental groups for very different types of josephson junctions .
all four experiments point towards an axion mass of 110 @xmath26ev .
further systematic measurements should still be performed to test whether these candidate signals are really due to axions .
the axion mass value of 0.11mev to which the various josephson experiments point to is compatible with current astrophysical and cosmological bounds on the axion mass .
99 g. bertone , nature 468 , 389 ( 2010 ) [ arxiv:1011.3532 ] r.d .
peccei , h. quinn , phys .
38 , 1440 ( 1977 ) p. sikivie , phys .
51 , 1415 ( 1983 ) f. wilczek , phys .
rev . lett .
58 , 1799 ( 1987 ) l.d .
duffy , k. van bibber , new j. phys . 11:105008 ( 2009 ) [ arxiv:0904.3346 ] p. sikivie , q. yang , phys .
103 , 111301 ( 2009 ) [ arxiv:0901.1106 ] c. arina , g. bertone , h. silverwood , phys .
d 88 , 013002 ( 2013 ) [ arxiv:1304.5119 ] r. bernabei et al .
phys . a 28 , 1330022 ( 2013 ) [ arxiv:1306.1411 ] j. hoskins et al . , phys
d 84 , 121302(r ) ( 2011 ) [ arxiv:1109.4128 ] m. arik et al .
( cast collaboration ) , phys .
107 , 261302 ( 2011 ) k. ehret et al .
( alps collaboration ) , phys .
b 689 , 149 ( 2010 ) [ arxiv:1004.1313 ] s.j .
asztalos et al .
( admx collaboration ) , phys .
104 , 041301 ( 2010 ) [ arxiv:0910.5914 ] p.w .
graham , s. rajendran , phys . rev .
d 88 , 035023 ( 2013 ) [ arxiv:1306.6088 ] c. beck , phys .
111 , 231801 ( 2013 ) [ arxiv:1309.3790 ] p. sikivie , n. sullivan , d.b .
tanner , phys .
112 , 131301 ( 2014 ) [ arxiv:1310.8545 ] r. li , j. wang , x - l .
qi , and s - c .
zhang , nature phys .
6 , 284 ( 2010 ) [ arviv:0908.1537 ] b.m
. roberts et al . , arxiv:1404.2723 m. tinkham , _ introduction to superconductivity _ , dover publ . , new york ( 2004 ) s. shapiro , phys .
11 , 80 ( 1963 ) c. hoffmann , f. lefloch , m. sanquer , b. pannetier , phys . rev .
b 70 , 180503(r ) ( 2004 ) [ arxiv : cond - mat/0409723 ] t.e .
golikova et al . , phys .
b 86 , 064416 ( 2012 ) [ arxiv:1202.5460 ] l. he , j. wang , m.h.w .
chan , arxiv:1107.0061 m - h .
bae , r.c .
dinsmore , m. sahu , h - j .
lee , a. bezryadin , phys . rev .
b 77 , 144501 ( 2008 ) [ arxiv:0802.2091 ] l. visinelli , p. gondolo , arxiv:1403.4594 p.a.r .
ade et al .
[ bicep2 collaboration ] , arxiv:1403.3985 p. sikivie , d.b .
tanner , k. van bibber , phys .
lett 98 , 172002 ( 2007 ) [ arxiv : hep - ph/0701198 ] l. visinelli , mod .
lett . a 28 , 35 ( 2013 ) [ arxiv:1401.0709 ] b.d .
josephson , phys .
lett . 1 , 251 ( 1962 ) c. beck , mod . phys .
lett . a 26 , 2841 ( 2011 ) [ arxiv:1110.5871 ] c. beck , physica c 473 , 21 ( 2012 ) [ arxiv:1008.2085 ] j.e .
kim , phys .
43 , 103 ( 1979 ) m.a .
shifman , a.i .
vainshtein , v.i .
zakharov , nucl .
b 166 , 493 ( 1980 ) m. dine , w. fischler , m. srednicki , phys . lett .
b 104 , 199 ( 1981 ) a.r .
zhitnitsky , sov .
phys . 31 , 260 ( 1980 )
m. weber , w. de boer , astron .
astrophys . 509:a25 ( 2010 ) [ arxiv:0910.4272 ] f. wilczek , arxiv:1401.4379 x - l .
xi , e. witten , and s - c .
zhang , phys .
b 87 , 134519 ( 2013 ) [ arxiv:1206.1407 ] g. rybka , a. wagner , arxiv:1403.3121 e. di valentino , e. giusarma , m. lattanzi , a. mechiorri , o. mena , arxiv:1405.1860 o. wantz , e.p.s .
shellard , phys .
d 82 , 123508 ( 2010 ) [ arxiv:0910.1066 ] g. raffelt , phys .
198 , 1 ( 1990 ) m. turner , phys .
197 , 678 ( 1990 ) | recently it has been proposed that dark matter axions from the galactic halo can produce a small shapiro step - like signal in josephson junctions whose josephson frequency resonates with the axion mass [ c. beck , prl 111 , 231801 ( 2013 ) ] . here
we show that the axion field equations in a voltage - driven josephson junction environment allow for a nontrivial solution where the axion - induced electrical current manifests itself as an oscillating supercurrent .
the linear change of phase associated with this nontrivial solution implies the formal existence of a large magnetic field in a tiny surface area of the weak link region of the junction which makes incoming axions decay into microwave photons .
we derive a condition for the design of josephson junction experiments so that they can act as optimum axion detectors .
four independent recent experiments are discussed in this context .
the observed shapiro step anomalies of all four experiments consistently point towards an axion mass of @xmath0ev .
this mass value is compatible with the recent bicep2 results and implies that peccei - quinn symmetry breaking was taking place after inflation . |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.